Documente Academic
Documente Profesional
Documente Cultură
Oxford Studies in
Political
Philosophy
Volume 1
Edited by
D AV I D S O B E L , P E T E R VA L L E N T Y N E ,
A N D S T E V E N WA L L
1
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
1
Great Clarendon Street, Oxford, OX2 6DP,
United Kingdom
Oxford University Press is a department of the University of Oxford.
It furthers the University’s objective of excellence in research, scholarship,
and education by publishing worldwide. Oxford is a registered trade mark of
Oxford University Press in the UK and in certain other countries
© the several contributors 2015
The moral rights of the authors have been asserted
First Edition published in 2015
Impression: 1
All rights reserved. No part of this publication may be reproduced, stored in
a retrieval system, or transmitted, in any form or by any means, without the
prior permission in writing of Oxford University Press, or as expressly permitted
by law, by licence or under terms agreed with the appropriate reprographics
rights organization. Enquiries concerning reproduction outside the scope of the
above should be sent to the Rights Department, Oxford University Press, at the
address above
You must not circulate this work in any other form
and you must impose this same condition on any acquirer
Published in the United States of America by Oxford University Press
198 Madison Avenue, New York, NY 10016, United States of America
British Library Cataloguing in Publication Data
Data available
Library of Congress Control Number: 2014953972
ISBN 978–0–19–966953–0 (Hbk.)
ISBN 978–0–19–966954–7 (Pbk.)
Printed and bound by
CPI Group (UK) Ltd, Croydon, CR0 4YY
Links to third party websites are provided by Oxford in good faith and
for information only. Oxford disclaims any responsibility for the materials
contained in any third party website referenced in this work.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
Acknowledgments
Contents
List of Contributors ix
Introduction 1
Steven Wall
viii Contents
Index 307
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
List of Contributors
Introduction
Steven Wall
2 Steven Wall
Introduction 3
4 Steven Wall
Suppose now that an actual state does have justified claims to govern in
the particular territory in which its members reside. Could it defend those
claims against unjustified aggression? Most people think so; but Frowe points
out in her “Can Reductive Individualists Allow Defense Against Political
Aggression?” that explaining why this is the case presents a challenge to any-
one who accepts a reductive individualist view of the morality of war.
(Reductive individualists hold that the rules governing killing in war are
equivalent to the rules that govern killing between individual people.) In
particular, Frowe discusses the argument of David Rodin that states do not
have the right to use lethal force to defend themselves against the unjust
aggression of another state if this aggressor state threatens only non-vital state
interests, such as the interests the state has in maintaining its jurisdiction
over its territory. An aggressor state could claim that its invasion will become
violent only if the victim state resists. The goods threatened by this type of
conditional force aggression, Rodin claims, are not weighty enough to justify
lethal self-defense by the victim state, at least on a reductive individualist
view. Frowe’s chapter presents a substantial response to Rodin’s challenge.
Among other things, she argues that non-vital interests, such as those impli-
cated in political rights to control a territory, when aggregated across enough
people do become weighty enough to justify proportionate lethal self-defense
and that once these rights are violated the threat to the victims changes sig-
nificantly, thus further justifying violence to defend them.
Next up is Eric Mack’s “Elbow Room for Rights,” which tackles a diffi-
cult problem for those who affirm a strong account of self-ownership. Self-
ownership rights theorists, such as Robert Nozick and Mack himself, hold
that people have very stringent rights over themselves and their legitimately
acquired extra-personal property. The problem with this view—a problem
pressed skillfully by Peter Railton and David Sobel in separate articles—is
that these rights appear to be so stringent that they implausibly restrict their
own exercise. For example, the rights seem to rule out very minor noncon-
sensual physical intrusions upon other right-holders.
But, if so, then right-holders may be “morally hog-tied” by the rights of
others, effectively disabled from exercising their own rights because doing
so would violate the rights of others. Mack agrees that this result is unac-
ceptable for the self-ownership view. In response, he presents an explanation
for the permissibility of minor physical intrusions by appealing to the idea
that the specification of the claim-rights of rights-holders must not system-
atically prevent them from exercising the liberty-rights that the claim-rights
purportedly protect. This “elbow room postulate,” Mack contends, allows
the self-ownership theorist to respond to the problem of minor intrusions
without downgrading the stringency of the rights that are central to the
self-ownership view.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
Introduction 5
Mack seeks to identify the rights, and limits to those rights, that people
have under favorable conditions. But might our rights change when cir-
cumstances are less favorable? This important and neglected issue is taken
up in Jonathan Quong and Rebecca Stone’s “Rules and Rights.” They pres-
ent a deontological approach to defining rights under non-ideal condi-
tions, one that differs from both remedial and rights-consequentialist
approaches to the problem. The key to their approach is to conceive of
rights violations as a cost or burden that should be distributed fairly among
free and equal persons. Rules that address this distributive problem can
ground individual rights that differ in content from the rights that people
would have under better conditions. In this way, Quong and Stone argue,
our rights change under non-ideal conditions, but they continue to reflect
the underlying values of freedom and equality that underwrite our rights
under ideal conditions.
While non-ideal conditions may affect the rights we have, they cer-
tainly create circumstances in which some people can take unfair advan-
tage of others. Thomas Christiano’s “What Is Wrongful Exploitation?”
presents a general account of unfair advantage taking. He argues that
wrongful exploitation, while wrong, is not a basic wrong. It piggybacks on
other wrongs. In particular, wrongful exploitation occurs when the
exploiter benefits at the expense of the exploited party by violating a duty
he owes to her. Christiano distinguishes his account from other accounts
that have been proposed and illustrates it with a number of examples. The
examples aim to bring out the key features of exploitative interaction, and
show that while it is parasitic on other wrongs it remains a distinctive
kind of wrong itself.
The final chapter by Ian Carter engages with fundamental issues of
method in political philosophy. “Value-freeness and Value-neutrality in the
Analysis of Political Concepts” asks what role does ethical evaluation play in
the analysis of central political concepts, such as power, freedom, and
democracy. Carter’s answer is subtle. He distinguishes value-freeness from
value-neutrality. The former refers to the absence of evaluative terms in the
definition of a political concept while the latter requires that one suspend
judgment on the merits of competing ethical positions when articulating a
political concept. (Neither of these ideas implies that one must detach an
analysis of a political concept from ethical concerns altogether.) Once prop-
erly distinguished, the notions of value-freeness and value-neutrality, Carter
argues, can both inform conceptual analysis of political concepts and help
to clarify our political ideas. In particular, value-free analyses are needed to
relate political concepts to the empirical world; and value-neutral analyses
are helpful in constructing normative theories on which different parties can
converge, allowing them to accept specific evaluations and prescriptions,
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
6 Steven Wall
PA RT I
DEMOCRACY
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
1
Justice: Social and Political
Philip Pettit
Every year, a number of different agencies rank the world’s cities in order of
livability, taking account of a variety of factors that affect people’s living
conditions.*1 The factors that bulk largest in the ratings are safety and crime,
urban design and environmental quality, public transportation, cultural
tolerance, business conditions, medical care, and so on. These elements all
belong in the domain of what government can provide or facilitate for
people but hardly bear at all on the democratic quality of the government
that ensures their delivery. Thus, the rankings treat people as relatively
passive beneficiaries of urban amenities and ignore the extent to which they
may act as active citizens to maintain control over what government does—
the extent to which they may combine to ensure that government operates
on terms that they generally endorse.
This imbalance of perspective in the international rankings of cities is
intelligible in light of what is allegedly one of the principal aims of the exer-
cise: to provide employers with information on how far employees can claim
hardship allowances for job relocation. Foreign employees need not have a
personal interest in the basis on which government provision for urban resi-
dents is determined, since they are not sufficiently permanent as residents to
be able to claim the democratic rights of citizens. Or at least they need not
have a personal interest in the mode of government decision-making so long
10 Philip Pettit
as the prospect of their being treated well remains bright or they retain a
ready right of exit in the event of that prospect’s fading.
When we think about the different ways in which our society may be
organized—the different institutional structures that may be established
there—we focus on the rival attractions of competing basic structures in a
way that parallels the focus on the rival attractions of the world’s cities. But in
this exercise it is vitally important that we look at those basic structures from
the perspective of active citizens as well as from the perspective of relatively
passive beneficiaries. It would be a serious mistake to enthuse about the way
our needs are satisfied under a certain institutional structure—say, a structure
in which a benevolent despot looks after our needs—without worrying about
how far we are in a position to combine with others to guard against any shift
in the will of the powerful, dictating the terms on which government operates.
This observation teaches an important lesson about what we should take
into account in thinking about the justice of a basic structure.1 In determining
whether a structure is just we have to think about whether it treats the people
who live under it—in particular, adult, able-minded, more or less permanent
residents2—as equals in what we take to be relevant respects; if we focused on
some other property of the structure we would not be thinking about its justice.
Thus, we have to take an impartial viewpoint that is capable of being shared
among the adult, able-minded, more or less permanent members of the
society, or at least among those who are ready to live on equal terms with
others (Pettit 2012; 2014).
The lesson of our observation in the case of livable cities is that in thinking
about the justice promised by a basic structure we ought to give consideration
to two aspects of the structure: first, the social justice between different indi-
viduals and groups that it would establish and, second, the political justice of
the relationship that it would set up between those individuals—roughly,
the citizenry—and the state or government that implements it. Social jus-
tice, as I shall use the phrase, is something that the state provides for its
people as the more or less passive beneficiaries of the system. Political justice
is something that it delivers for its people in their role as active citizens: that
is, as members of the society who ought presumptively to share in setting
the terms on which their state acts. In what follows, I shall assume that the
citizenry are identified appropriately and that they are the adult, able-
minded, more or less permanent residents of the society.
1
I ignore here the issue as to whether justice is properly restricted to the basic struc-
ture of a society without attention to the justice exhibited by individuals acting under
that structure. For a critique of the restriction, see Cohen (1989).
2
Justice imposes special requirements on how those who are not adult, not able-
minded, or not permanent residents should be treated but I ignore those in the present
context. I am concerned with general issues of domestic justice as distinct from special
issues of domestic justice or, indeed, issues of international justice.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
3
I consider political justice under a broad conception of what it involves in Pettit
(2012; 2014). Under this conception it coincides with what is often described as the
legitimacy as distinct from the justice—I would say, the social justice—of the structure;
see Simmons (1999).
4
There are exceptions, of course. Thus, Juergen Habermas (1995) has always insisted
on their connection as does Rainer Forst (2002), who operates within broadly the same
tradition. Charles Beitz (1989) and Thomas Christiano (2008) both focus on the impor-
tance of equality in democracy and as a result they also tend to maintain a connection
between concerns of social and political justice.
5
But while Rawls (1971) suggests that the political liberties are subordinate in an
instrumental role, he stresses their intrinsic importance in shaping the character of citi-
zens. And in his later reply to Habermas, Rawls (1995, 163) goes further still in their
defense, implying in the terminology of his interlocutor that the political liberties are
“co-original and of equal weight” with other basic liberties. I am grateful to Paul Patton
and Steven Wall for alerting me to these aspects of Rawls’s position. See too n 9. For a
congenial assessment of how far Rawls can take this later line and preserve his theory of
justice as fairness, see Wall (2006).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
12 Philip Pettit
to one another. It should be clear that, like anyone else’s, your ideal of polit-
ical justice—say, your rich democratic ideal of political justice—is liable,
should it be realized, not to support your ideal of social justice: other people
may not combine democratically to support it. And that possibility raises
some serious questions.
Should you be prepared to pay a cost in one form of justice for a benefit
in the other? And if so, how should the trade-off go? Should you be ready,
at the limit, to drop all demands of political justice for the realization of
social? Or should you be ready to drop all demands of social justice for the
realization of political?6
I argue that under some fairly standard assumptions there is a powerful
case for giving a certain priority to political. The very activity of proposing
an ideal or theory of social justice presupposes under those assumptions that
there ought to be a politically just, suitably democratic procedure in place
in the society targeted. And while the content of an ideal or theory of polit-
ical justice may require a degree of social justice in the targeted society—we
shall see in the conclusion that any plausible theory will require this—the
activity of proposing such an ideal is not tied in the same way to a presup-
position of social justice.
In the following three sections I defend three progressively stronger the-
ses about the priority that political justice ought to enjoy under standard
assumptions and then I conclude in section 5 with a brief comment on the
theoretical and practical implications. The theses that I defend in the fol-
lowing three sections hold that as theorists of social justice we are commit-
ted by assumptions almost all of us share to holding that in any society for
which we prescribe a scheme of social justice:
• some politically just process of decision-making ought to be established;
• this process ought to require a form of democratic approval; and
• this process ought to introduce a form of democratic control.7
6
While these issues are often neglected among political philosophers, Laura Valentini
(2013) is a notable exception. I find her views broadly congenial and I was stimulated in
my own thinking about the issues by a presentation that she made in a seminar at
Princeton in a debate with Ryan Davis.
7
The argument of the chapter takes us from assumptions we allegedly share as theo-
rists of social justice to commitments that they entail about the importance and, indeed,
priority of political justice. The argument skirts issues of feasibility, since it explores a
connection between our views of social justice and our views of political rather than
directly supporting any particular institutional recommendations. But it raises a related
question as to what we ought to think if we judge that the commitments in political
justice that we make as theorists of social justice are so demanding as to be unrealizable.
The lesson would seem to be that we should give up on some of the starting assumptions
about social justice but I do not explore that issue in the chapter.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
Few if any contemporary thinkers who take social justice seriously hold by
their particular view of justice in the manner in which the protagonists in
the European wars of religion held by their different positions. However
fully and fervently they believe in their particular vision of social justice,
they shrink from maintaining that it ought to be imposed by the coercive
state without regard to whether or not others endorse it. And as they shrink
from investing their own view of social justice with such a special, peremptory
status, of course, so they deny that status to the views of others. They agree
in each renouncing any form of crusading vanguardism or sectarianism in
favor of an attitude of mutual respect or forbearance.
This disposition of forbearance is unsurprising in view of an assumption,
shared by theorists on all sides, about the more or less egalitarian character
of justice and about the more or less egalitarian character of their rival the-
ories. It is axiomatic for all, and manifestly axiomatic for all, that a just basic
structure must treat citizens as equals in their relations with one another. If
the defenders of rival theories are forbearing in their attitudes towards one
another, that is because they assume that each of them is seeking an inter-
pretation of what such treatment requires (Dworkin 1978). Thus, there is
no suggestion that they would be forbearing in relation to a social theory—
it could scarcely be called a theory of justice—that dismissed out of hand,
or clearly downgraded, the claims of some.
The disposition of forbearance among theorists of social justice may be
grounded in any of a variety of considerations: for example, that anything
short of this disposition is a recipe for civil war; that no system of social justice
will be stable if it is imposed on people without regard for their attitudes; that
under any plausible view of social justice respect for the views of others is itself
an essential requirement; that the best scheme of social justice is likely in any
case to be a compromise between rival theories; and so on. I assume that for
one or a number of such reasons, every contemporary theory of justice presents
itself in the forum of public debate as a theory offered for the consideration of
others, not as a blueprint to be coercively implemented, regardless of people’s
attitudes towards it. And I assume that this is as it should be: that the consid-
erations rehearsed do indeed support such a disposition of forbearance.8
This disposition of forbearance is certain to be triggered in real-world
scenarios. For it is a commonplace that in almost every society there are
8
This is to favor the democratic line in resolving Richard Wollheim’s (1962) paradox
in the theory of democracy.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
14 Philip Pettit
9
John Rawls (2007, 2) might be taken to endorse the line presented here when he
writes: “Political philosophy can only mean the tradition of political philosophy; and in
a democracy this tradition is always the joint work of writers and of their readers. This
work is joint, since it is writers and their readers together who produce and cherish works
of political philosophy over time and it is always up to voters to decide whether to
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
The commitment to the idea that there ought to be a just process of deci-
sion-making for accommodating their differences is inescapable for theories
of social justice. If the forbearing defenders of different theories were to reject
the commitment, then they would have to embrace one of two unpalatable
alternatives. The first would be to reject any semblance of realism about
politics and argue that short of theoretical convergence, there should be no
practical accommodation among different views of social justice; the debate
should continue indefinitely. The second alternative would amount to
rejecting any suggestion of idealism in politics; it would accept that, no just
accommodation of differences being possible, each society is entitled to let
non-normative forces shape how it goes, indifferent to people’s rival views
of social justice.
To accept either of these views would be to take the theory of social jus-
tice to be irrelevant to public life. The first view would restrict normative
theorizing about social justice to the seminar room, which is the only loca-
tion that welcomes continuing, possibly endless debate. The second view
would despair of the capacity of normative theorizing to have any impact
on the life of a society. To endorse one or the other view would be to deny
any practical point to the theory of social justice, insulating normative
thinking about politics from the political world. Assuming that theories of
social justice can scarcely support this attitude—it would deprive them of
any application, after all—I conclude that they must commit to the idea
that each society ought to have a just process of decision-making to estab-
lish compromises between the rival views that emerge there.
The commitment ascribed here to theories of social justice is existential
in character, not substantive. It involves accepting, not that a particular
decision-making process should be established in the society for which
rival theories make recommendations, but that one or another just decision-
making process—maybe this, maybe that—ought to be put in place. For all
that we have said so far, theories of social justice may be silent on the form
that this process should take; on that matter they will presumably defer to
the theory of political justice. But as we shall see in the two following sec-
tions, there are other assumptions built into theories of social justice that do
put constraints on what form a theory of political justice ought to assume.
This line of argument suggests, then, that all forbearing theories of social
justice are committed to thinking that in circumstances of disagreement some
just process of decision-making—to be identified, presumably, in the theory
of political justice—ought to be introduced to accommodate their differences
in advance of whatever theoretical convergence may be expected. Each theory
embody their ideas in basic institutions.” Thanks to Ben McKean and Larry Udell for
drawing my attention to this passage.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
16 Philip Pettit
We have argued that elaborate theories of political justice, and all theories of
social justice, are committed under the assumption of mutual forbearance
to holding that some just process of decision-making between rival proposals
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
18 Philip Pettit
pragmatic message that you are someone who can understand, assess, and
act on the recommendation. Communicating both messages at once, my
speech act amounts to an invitation to you to do X on the basis of the rea-
sons presumptively supporting it; the suggestion is that you are able, enti-
tled, and even perhaps obliged to take that line.
The pragmatic message in a recommendation is not capable of being
cancelled. I can hardly say: I recommend that you do X but I do not mean
to suggest that you are capable of appreciating what a recommendation is.10
And so the presence of the pragmatic message puts constraints on the
semantic content that the recommendation can carry. I may recommend
that you should do X, where I can and do assume—although perhaps mis-
takenly—that X is within your deliberative control: something that you
can do or not do, depending on your assessment of the pros and cons. But
I cannot recommend responses that do not fit this constraint. I cannot rec-
ommend, for example, that you should blush or fall in love or be intimi-
dated, where it is manifest to all that those responses lie outside your
deliberative control.11
The fact that each theory of social justice makes a recommendation or set
of recommendations rather than issuing a series of directives or commands
carries a clear pragmatic message. It presupposes that an addressee of the
theory has the capacity to understand what is recommended, to assess the
recommendation on the basis of the pros and cons, and to try to act on it in
the event of a positive assessment. And it communicates that presupposi-
tion to addressees, implicitly inviting them to exercise the presupposed
capacities in relation to the recommended course of action. Whatever the
semantic content of the utterance—whatever the response recommended—
this pragmatic message is absolutely unambiguous.
But the recommendation or set of recommendations made by any theory
of social justice is quite distinctive in character, since it is offered in public
10
On the difference between pragmatic messages that can be cancelled and those that
cannot, see Jackson (1987).
11
On the assumption that it is manifest that you cannot deliberatively control blush-
ing or falling in love or being intimidated, such a recommendation would involve a
pragmatic inconsistency, as it is called: an inconsistency between the pragmatic and seman-
tic message of the utterance. The paradigm of such an inconsistency, due to G. E. Moore,
is “p but I don’t believe it.” This utterance communicates pragmatically that I hold by the
belief that p but communicates semantically that I do not believe that p. There is no
inconsistency in its being true that p and that I do not believe it, as there is no inconsist-
ency in my recommending that you X and its being impossible for you to X. But there is
an inconsistency in my implicitly communicating that I believe that p while claiming
explicitly that I do not believe it, as there is an inconsistency in my implicitly communi-
cating that you can deliberatively control X-ing when it is manifest to all that X-ing lies
beyond your control.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
to all the citizens of any society that falls within its purview; this society may
or may not be the theorist’s own. If a theory were to avow the aim of offer-
ing advice only to an administrative elite, as Renaissance political tracts
offered advice only to princes, then by current standards it would under-
mine its own credentials and fail to count as a theory of justice proper
(Habermas 1971). Indeed it would even run into conflict with the medieval
principle according to which what affects all ought to be debated and
approved by all: Quod omnes tanget ab omnibus tratactari et approbari debet
(Congar 1958).
The publicity of the recommendation offered impacts on the pragmatic
message of the theory and means that it communicates a very specific form
of invitation to addressees. Each theory communicates the message that all
the citizens it addresses can appreciate and try to act on the sort of recom-
mendation that it puts before them. And because it is offered as a public
or manifest matter—that is, in such a way that each knows that each is
addressed, each knows that each knows this, and so on—it communicates
the message that they can appreciate and try to act on it jointly. Thus each
theory communicates the message that it is within the competence of the
public or citizenry as a whole to collectively assess and act on the recommen-
dations made: to establish the institutional means whereby those recom-
mendations can be considered and, if approved, implemented. It amounts,
in effect, to an invitation to the citizenry to get together and determine a
shared view of the proposal offered.
We saw earlier that each theory of social justice is committed to the idea
that under the conditions of disagreement that obtain in every society, there
ought to be a just decision-making process in place for establishing a compro-
mise between competing views. Once we recognize the pragmatic message
that each theory communicates in virtue of making public recommendations
to all the citizens of a relevant society, we can see that this commitment to
the need for a just decision-making process has a more determinate charac-
ter than registered earlier. It amounts to a commitment to the need for a just
decision-making process in which each citizen, each member of the public,
is invited to participate in a deliberative manner.
By this account, then, the commitment made in a theory of social jus-
tice is that each society ought not only to put in place a just process for
accommodating differences between competing theories but that it ought
to put in place a process that all the members of the society are ready to
accept. This requirement strengthens the earlier requirement that the pro-
cess for accommodating differences should be just. The requirement it puts
in place—that all be ready, presumably with good reason, to accept the
process—entails that it is just, as the earlier claim maintained, but not the
other way around.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
20 Philip Pettit
12
The distributive character of the system, as envisaged here, allows for distribution in
two dimensions: first, in the processes employed in different domains of decision-making
and, second, in the agents—the representatives of the people, as it is natural to say—who
operate those processes.
13
Any veto will deny the person exercising it access to a shared process—and pre-
sumptively a shared benefit—as it will deny this to others too. I think of that cost as
general rather than special. While it may induce someone to settle for less than what they
consider best, it is very different from the special cost associated, for example, with being
bullied not to cast a veto or ostracized as a result of doing so. I assume that if the shared
benefit of the arrangement is sufficient to get people to accept a process, then that process
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
must promise to serve them reasonably well. I set aside the problems that strategic
game-playing might raise.
14
This virtual mode in which a group endorses a process is logically prior to any
active mode of endorsing it: say, the way members endorse a process when they vote
unanimously in favor. Even if members vote unanimously in favor of a process, this will
carry weight only insofar as they virtually endorse the unanimitarian process itself. And
if they seek to endorse that unanimitarian process in some independent voting procedure,
then any support they provide will carry weight only because of the virtual acceptance
of that procedure. The lesson is that there is no collective endorsement of any proposal,
including the proposal to follow a certain process, without the purely virtual endorse-
ment of some process of endorsement. It parallels the lesson that Lewis Carroll (1895)
taught us in the case of reasoning: that there is no way in which an individual can derive
conclusions from premises without the purely virtual endorsement of some principle or
rule of inference.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
22 Philip Pettit
24 Philip Pettit
life—regular meals, a roof over your head, and so on—and getting your
preferences to shift appropriately.
Freedom, by this argument, is robustly demanding insofar as it requires
that if you are free in a choice between certain options like living inside or
outside prison then those options must each be open doors. You must be
able to choose according to your actual wishes and it must be the case that
had your wishes been otherwise you would also have been able to choose
according to those wishes; you must enjoy non-interference in either event.
But this conception of freedom is not the only one that makes it into a
robustly demanding good. A more traditional construal holds that freedom
is robustly demanding in another measure also.
On this traditional reading, freedom in a choice requires that your being
able to choose according to your wishes, regardless of what you wish, should
not be dependent on anyone else allowing you that license. If you were
dependent on the goodwill of another for being able to choose according to
your own will, then the other’s will would be in ultimate charge, not your
will; and in that sense you would be unfree (Pettit 2014).15 In order to enjoy
a properly free choice, the options must not only be open doors; there must
be no doorkeeper on whom you depend for leaving them open. You must
be able to choose as you wish, not only regardless of what you want to
choose, but also regardless of what others want you to choose or be able
to choose.
As it is with freedom, so it is with any form of justice. Take the private
justice, first of all, that you may experience in your dealings with me. You
do not enjoy justice at my hands if I merely happen, as luck would have it,
to satisfy the actual claims that you make on me; I must be disposed to
satisfy those claims robustly across variations in my particular inclinations.
I can act justly towards you in the absence of such a disposition—I can
satisfy your claim in justice—but without that disposition I do not count
as a just person in my interaction with you: I do not give you the benefit
of justice.
Now consider the justice, social or political, that you may hope to expe-
rience in your relations with your community as a whole. As in the other
case, you will not enjoy this sort of justice—this public as distinct from
private justice—if it is merely good luck that you are given satisfaction of
your claims, whatever they are taken to be. But neither will you enjoy this
sort of justice if that depends on others in your society being suitably virtu-
ous: that is, being disposed to grant you satisfaction of your public claims,
as they may be disposed to grant you satisfaction of your private. It must be
15
This construal is associated, in my view, with the long republican tradition. (See, for
example, Pettit 1997; Skinner 1998; Pettit 2012.)
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
26 Philip Pettit
16
This observation means that I dissent from the position G. A. Cohen (2008, 269)
gestures at when he sets up a contrast between the “fundamental principles of justice”
with which the philosopher ought to be concerned and the “rules of social regulation,”
which bear on the presumably empirical matter of how they are best implemented in
actual circumstances. Justice requires rules of social regulation as an inherent matter, on
the account given here, not just as instruments for realizing prescribed patterns. It is
those rules or institutions that are going to determine whether or not the patterns that
our principles of justice identify are robustly in place. And unless they are robustly in
place, the realization of those patterns does not provide justice. In this criticism of
Cohen’s view, I join with Charles Larmore (2012), and Jeremy Waldron (2013), though
for somewhat different reasons.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
28 Philip Pettit
The only basis for questioning the need for democracy in this sense is the
thought that the process for accommodating differences between theories of
social justice might be taken out of human hands altogether, or at least out
of the hands of intentional designers. As a matter of logical possibility, an
impersonal or non-intentional process might establish a compromise scheme
of social justice that gave each the same claims against others and gave them
on a basis that was robust over shifts in individual inclination or bias. But
there is no real-world possibility or prospect of such a robotic process,
although variations on the idea have beguiled a number of thinkers. Perhaps
it was a fascination with this idea, for example, that explains why Frederic
Hayek (1988) looked for a system of decentralized, common law courts—
surely a fantasy on stilts—that would generate a bare form of social justice
as by an invisible, unintentional hand.
This line of argument leads me to support the third of my progressively
stronger theses. Theorists of social justice are not only committed to the
idea that there ought to be a process for accommodating differences between
rival theories that is just and has democratic approval. As advocates of jus-
tice, robustly understood, they are also committed to the claim that the
process ought to embody a degree of democratic control, giving people an
equal voice in exercising control over how that process goes. This constraint
is stronger again than the previous two. A process that gave people some-
thing like an equal voice, satisfying this requirement, would have to satisfy
the earlier requirements too but not the other way around. It would pre-
sumably attract the approval of all—or at least all who are willing to live on
equal terms with others—and, treating all as equals in that respect, it would
count moreover as just.
As with the argument for the first and second constraints, this argument
about the commitment of theories of social justice applies also to elaborate
theories of political justice. They too are committed, as we have seen, to the
idea that there ought to be a just process for accommodating differences
between rival theories and, in particular, a process that attracts everyone’s
approval. But because they are theories of justice, albeit political rather than
social justice, the decision-making process that they require has to involve a
mode of framing and imposing law that gives each an equal voice and
thereby ensures that the satisfaction of people’s claims of political justice is
suitably robust.
The reference to theories of political rather than social justice prompts a
final observation that bears on the sort of democratic control that theorists
of robust social justice must require. If the political process were subject to
the say-so of a particular individual or dynasty or elite, as we noticed, then
it could certainly not deliver social justice with the required robustness.
Such a failure of political robustness would also affect social robustness,
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
17
For an insightful discussion of democracy as a robustly demanding value, see
Southwood (2014). Where I invoke the eyeball test here for testing both social and polit-
ical justice, I invoke the related tough-luck test for checking political justice in Pettit
(2012, 2014), restricting the eyeball test to social justice.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
30 Philip Pettit
electorate, that acts in the name of the people. There are many different
ways in which that specification might be institutionally implemented, of
course, and the process has to be able to accommodate differences between
these elaborated proposals. But the specification is still sufficiently demand-
ing to put many theories of political justice off the shortlist of plausible
candidates. It holds out the prospect that only a small family of models
may be available to exemplify what a basic process of political justice might
look like.
32 Philip Pettit
Bibliography
Aquinas, T. (1958). Summa Theologica. Madrid, Biblioteca de Autores Cristianos.
Beitz, C. (1989). Political Equality: An Essay in Democratic Theory. Princeton, NJ,
Princeton University Press.
Berlin, I. (1969). Four Essays on Liberty. Oxford, Oxford University Press.
Carroll, L. (1895). “What the Tortoise said to Achilles.” Mind 4: 278–80.
Christiano, T. (2008). The Constitution of Equality: Democratic Authority and its
Limits. Oxford, Oxford University Press.
Cohen, G. A. (1989). “On the Currency of Egalitarian Justice.” Ethics 99: 906–44.
Cohen, G. A. (2008). Rescuing Justice and Equality. Cambridge, MA, Harvard
University Press.
Congar, M.-J. (1958). “Quod Omnes Tangit ab Omnibus Tractari et Approbari
Debet.” Revue Historique de Droit Francais et Etranger 36: 210–59.
Dworkin, R. (1978). Taking Rights Seriously. London, Duckworth.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
34 Philip Pettit
Wall, S. (2006). “Rawls and the Status of Political Liberty.” Pacific Philosophical
Quarterly 87: 245–70.
Walzer, M. (1981). “Philosophy and Democracy.” Political Theory 9: 379–99.
Watson, A. (1985). The Digest of Justinian, Four Volumes. Philadelphia, University of
Pennsylvania Press.
Wollheim, R. (1962). A Paradox in the Theory of Democracy. Philosophy, Politics and
Society; Second Series. P. Laslett and W. G. Runciman. Cambridge, Cambridge
University Press: 71–87.
OUP CORRECTED PROOF – FINAL, 26/03/15, SPi
2
Voting and Causal Responsibility
Geoffrey Brennan and Geoffrey Sayre-McCord
1. INTRODUCTION
1
There is a second-order complication that we should dispose of at the outset: J might
win the election following an exact tie among all voters. The decision over who wins in the
event of a tie might be determined by tossing a coin; and J, rather than K, might win that
coin toss. Each voter, in these cases, is such that had she abstained, or voted differently, she
would have brought about the victory of K. And in the case where J wins by exactly one
vote, if a voter had abstained, she might have brought about the victory of K via a tie-break-
ing mechanism. So, if voter A abstains rather than votes for J, in the first case, she would
cause K to win by a majority of 1; or, in the second case, she would give rise to a situation
in which K might win via a tie-breaking procedure. What this means is that the idea of
being pivotal is slightly different depending on whether the option under consideration is
voting for an alternative candidate or abstaining, and on whether the number of voters is
odd or even. Here, we shall treat the idea of being pivotal capaciously—allowing it to cover
all cases in which A’s voting behavior could influence which candidate wins.
OUP CORRECTED PROOF – FINAL, 26/03/15, SPi
the extent to which decisions about which candidate to vote for will
be influenced by the voter’s individual self-interest;
and whether it will be rational to vote at all, given the stakes involved
in the election for the individual as she perceives them, and the
expected closeness of the race and the alternative actions to voting
available.2
In most large elections, clearly, h (even capaciously defined in the sense of
footnote 1) will be miniscule. Given this fact,3 rationality considerations (as
RCV understands them) can explain:
1. why individual voters do not expend much effort in acquiring rele-
vant information about candidates and their policies. After all, their
votes are very unlikely to make a difference to the outcome. (This is
the basis of Downs’ (1957) claims about electoral “rational ignorance.”
For a more recent treatment, see Caplan (2007).)
2. why considerations of individual interest as opposed to public
interest are likely to play a smaller role in decisions about how to vote
than in those same agents’ market decisions. After all, since the chance
that one’s voting decision will have an impact on the outcome is far
lower than in market contexts, where one’s choice is usually decisive,
the chance that it will have an impact on one’s individual interest is
correspondingly small. (This is the central upshot of the “expressive”
account of voting behavior—as analyzed, for example, by Brennan
and Lomasky (1993).)
3. why certain factors, and not others, are likely to influence aggregate
levels of voter turnout.4 After all, if the chance that one’s vote will
make a difference to the outcome is exceedingly small, the incentive
provided by the thought that one’s vote will make a difference will be
small as well.
2
It is perhaps worth emphasizing that the question of whether it is rational to vote or
not involves a deployment of the logic of rationality that is somewhat at odds with its use
in “rational choice theory.” Charges of irrationality involve argument from presumed
desires and beliefs on the one hand to actions that are inconsistent with those desires and
beliefs on the other. The rational choice theory approach involves argument from
observed behavior “backwards” to the beliefs and desires that are thought to motivate
that behavior. Here, rationality works as an analytic assumption rather than an issue to
be determined in any specific case. So, the presumption in RCV is (or ought to be) that
both voters and abstainers are rational: the question then becomes—what differences in
beliefs and desires of the two sets of agents would explain these behavioral differences?
The notion that one of these groups might be ‘rational’ while the other group is not is
alien from the spirit of the RCV enterprise—a fact which did not prevent some early
RCV theorists from speculating as to whether voting is “rational” or not. See for example,
Tullock (1967).
3
For example, in US Presidential elections, at most of the order of 1/12500 and
almost certainly rather smaller than this. See Brennan and Lomasky (1993) chapter 4.
4
The classic treatment is Riker and Ordeshook (1968).
OUP CORRECTED PROOF – FINAL, 26/03/15, SPi
5
Of course, expected benefit might be a reason for voting if the stakes for the voter
are huge—but then it is the size of the stakes rather than the probability of making a
difference to the outcome that is the primary “reason” to vote.
6
The variations are far from minor, but they share enough, we think, for it to make
sense to consider them together, not least their rejection of the idea that what matters is
that one’s vote will be pivotal.
OUP CORRECTED PROOF – FINAL, 26/03/15, SPi
cient for it.7 To take an example from Tuck, a police officer will have killed
a robber—will be the cause of the robber’s death—even if some other officer
would have shot the robber had the first missed. So the first’s accurate
shooting was not necessary in order for the robber to die. And a lifeguard
saves your live—is the cause of you not drowning—even if, had she not
saved you, another lifeguard standing by would have: so the first lifeguard’s
action was the cause of your being saved even though her saving you was
not necessary for you to survive. Similarly, on their view, one’s vote can be
a cause of an electoral outcome even in cases where there are more than
enough votes for victory, so even when one’s voting is not necessary for the
outcome (although Goldman and Tuck differ concerning just when one’s
vote is properly seen as a cause in such cases). As a result, they argue, h is not
the probability that deserves attention in thinking about whether one’s vote
will be causally efficacious; rather, what is important is the ex ante probabil-
ity that one’s vote will be among those that are, in the appropriate way,
sufficient for electoral victory. (In what follows, we refer to this as the
“Goldman/Tuck account.”)
So, while RCV and Goldman/Tuck agree that in thinking about whether
to vote a central question is whether in voting one will bring about, or
cause, the desired outcome, they differ concerning their understanding of
what it takes to be a cause. On the RCV view, the relevant test is counter-
factual: would the outcome have been the same even if you had voted dif-
ferently. If yes, then your vote is not pivotal, and so not (on this view)
causally efficacious. On the Goldman/Tuck account, even if your vote is not
pivotal, it will be causally efficacious if it is among the votes that are, in the
appropriate way, sufficient for the outcome.
There are some important differences between Goldman and Tuck con-
cerning just which sufficient set of votes matters in the context of voting. In
particular, on Tuck’s view, it is whichever minimally sufficient set actually
settled the election. The idea shows up nicely, as Tuck emphasizes, in a roll-
call vote, where at a certain point victory is secured by someone’s vote,
together with those before, even if those following in the roll call would
have voted in the same way.8 The same idea shows up, although with com-
plications due to the electoral college, in thinking about the causal efficacy
of votes cast on the West Coast or in Hawaii in US presidential elections.
Often, the outcome is settled, by the votes cast earlier, farther East, in a way
7
We write of being sufficient “in the appropriate way” because Goldman and Tuck
differ when it comes to identifying which votes are a part of what is sufficient for the
outcome. We highlight the difference below.
8
Tuck, p. 51.
OUP CORRECTED PROOF – FINAL, 26/03/15, SPi
that makes the later votes immaterial to the actual outcome.9 For other
votes as well, Tuck maintains, there is some subset of the votes cast for the
victor—a minimally sufficient set—that will have been causally efficacious,
because actually sufficient for the outcome.10 On Tuck’s reckoning, the ex
ante probability that one’s vote for a victorious candidate will be causally
efficacious—the probability that it will be in the minimally sufficient set—
is normally much higher than h, even in large landslide elections. For
instance, in an election with a million voters, where the victor is expected to
win with 60% of the vote, the probability that one will be in the minimally
sufficient set, g, is 0.66 (whereas h is vanishingly small). And this is true
even though, as is often the case, we cannot know which subset of the votes
actually proved to be the sufficient set.
Suppose in a two candidate race, J emerges as victor, receiving N votes
out of the total voting population of M. Now, since (M – N) voters voted
for the alternative (K), the minimal set of voters J required for victory was
in fact (M – N + 1). We do not know which particular voters compose that
minimal set. But Tuck supposes, for this example, that the votes were
counted in a particular order and there is indeed a determinate set of voters
who compose that minimal set. So there is a fact of the matter about which
voters these were. A, who voted for J, can reasonably ask what the probabil-
ity is that she (A) herself was a member of that minimally sufficient set.
And that probability is just the number of voters required for the mini-
mal majority divided by the actual number of J-voters—or (M – N + 1)/N.
Denote this probability, g. Tuck sees g as the appropriate measure of the
probability that a voter is in the relevant minimally sufficient set, and thus
a cause of the victory.
To take a simple arithmetic example, J gets 6000 votes out of a total of
9999 voters. So 3999 individuals voted for K. For J to win, J would have
needed 4000 votes: this is the minimal set of voters required to bring about
9
Tuck’s account faces complications, of course, in cases where votes are cast simulta-
neously, since the roll-call voting method of isolating which set of votes was, in fact,
sufficient, depends on temporal order. This fact plays an important role in motivating
Goldman’s account, which takes simultaneous voting as the normal case (on the grounds
that our voting conventions treat the timing of ballots, and of their being counted, as
irrelevant (p. 213)).
10
Writing about British parliamentary elections, Tuck notes that “at some point in the
course of the evening one candidate’s piles of ballot papers add up to the precise figure
necessary for a majority, and those ballot papers have therefore by themselves accom-
plished the task of electing him.” And he goes on to note that in the event of a recount,
the original count proves not to have settled the election and “the set which first reaches
the majority in the second count will then do so, or in the third count, or in however
many counts are necessary to satisfy the returning officer and the candidates that a proper
count has been made” (p. 43).
OUP CORRECTED PROOF – FINAL, 26/03/15, SPi
J’s victory. So each of the 6000 actual J voters has a two-thirds chance of
being in that set—and should on this basis be seen as having a two-thirds
chance of being causally responsible for J’s victory.
Goldman, in contrast, holds that in normal elections a vote is a “partial”
or a “contributory” cause as long as it is a member of any minimally sufficient
set, privileging no particular set as the one that, in the event, was s ufficient.
This means that every vote cast for a victor, even in a landslide, counts as
causally efficacious, not just those that are in some privileged minimally suf-
ficient set.11 As a result, the ex ante probability that one’s vote for a victorious
candidate will be causally efficacious, as Goldman thinks of things, is 1. (Of
course, when one does not know for certain which candidate will win, the
ex ante probability that one’s vote will be causally efficacious in bringing
about victory will be less than 1, reflecting the probability, whatever it is,
that another candidate will win, but in any case, still much higher than h.)
Importantly, in developing his view, Goldman focuses on cases of what
might be called simultaneous over-determination, where the sets of suffi-
cient conditions are all in place at the same time. In these cases it is hard, to
say the least, to make sense of the idea (important to Tuck) that one of the
minimally sufficient conditions takes precedence over all of the others. Yet,
Goldman recognizes the challenge posed by cases of non-simultaneous
over-determination (e.g. the roll-call vote and our national elections), where
the order of the votes, or the order in which they are counted, suggest that,
as a matter of fact, one sufficient condition pre-empts all the others. In
response, Goldman argues that in many elections we have a conventional
system in place that “abstracts from this actual or ‘natural’ order and consid-
ers all the votes on an equal basis” turning an apparent case of pre-emption
into one of simultaneous over-determination.12 We are unsure as to whether
the difference just noted is simply due to Tuck’s focusing on cases of non-
simultaneous over-determination, where one sufficient set of voters is rea-
sonably seen as pre-empting the others, and Goldman focusing on cases
of simultaneous over-determination, where there seems to be no grounds
for privileging one set of sufficient conditions over others.
11
Goldman advances this idea as a way of understanding Mackie’s INUS account of
causation in light of which something is a partial or contributory cause if and only if it is
“an insufficient but necessary part of a condition that is itself unnecessary but sufficient for
the result” (p. 206).
12
Goldman points out that “In the United States House or Senate, for example, a
roll-call vote is completed even if the outcome is clear long before the last vote has been
voiced. This is because, officially, votes are not counted or ‘registered’ until all have been
voiced. Because of this conventional feature, the causal impact of a late vote is not really
preempted by a collection of early votes. From the official, conventional perspective, they
are all simultaneous; hence, their causal statuses are perfectly symmetric” (p. 213).
OUP CORRECTED PROOF – FINAL, 26/03/15, SPi
It may be that each would accept the views of the other for the relevant
cases. What is important for us here is that Goldman and Tuck both reject
RCV, along with its emphasis on h, and recommend thinking of causation
in terms of the sufficient, rather than necessary, conditions for an outcome.
For the most part, differences between Goldman and Tuck (to the extent
there are differences) won’t matter to what follows, though where it does we
will default to Tuck, using g (interpreted as his view, rather than Goldman’s,
would recommend). This is for two reasons. The first is that we think that
in many cases (in elections and otherwise) the kind of convention Goldman
relies on in forestalling worries about pre-emption is not in place. And we
take it that when it is not we should be thinking in terms of pre-emption (as
per Tuck) rather than simultaneous over-determination. The second is that
the interpretation Tuck’s account gives to g makes that probability more
directly comparable to RCV’s h than is the interpretation provided by
Goldman’s account (which would set the probability at 1 in cases where one
is certain which candidate will win).
The contrast between g and h is striking. For a start, g is very much larger
in all democratically relevant cases. As noted in footnote 3, the probability
of an exact tie (h) between candidates in a US presidential election too close
to call is estimated at around 1/12,500. But for just such an election, g will
be quite close to 1. Even if the margin turned out to be 4 million voters out
of 120 million (perhaps not so very close after all) then g is 58/62 or roughly
0.94. On Tuck’s view, this probability measures the probability that one will
be a cause of J’s victory.
Note too that g and h vary in different ways. As electorate size increases,
the probability of an exact tie declines roughly with the square root of size:
so when M doubles, h declines by about 40%. By contrast, g is a matter of
relevant proportion, increasing as the size of the minimally sufficient set
increases.
It is worth emphasizing that g can be thought of as an ex post probability—
to be calculated after the election is settled—whereas h cannot be. After the
election, you will know whether you were pivotal or not. But the decision
as to whether to vote or not depends on the probability that one will be a
cause of the outcome in question; and so g should properly be thought of
here in ex ante expected terms. The ex ante expected value of g will be based
on the expected size of the majority. The voter properly expects h to be tiny;
but g is likely to be substantial.
In developing their accounts of how it is our votes might be causes of an
electoral victory without being required for it, Goldman and Tuck are espe-
cially interested in why voters “should vote.” They want to identify reasons
people have to vote even in cases in which one candidate will predictably
beat others, with or without their vote. Goldman explicitly recognizes that
OUP CORRECTED PROOF – FINAL, 26/03/15, SPi
the reasons in play might be either prudential or moral; and though his
own account focuses on the moral aspect, he thinks his reasoning has clear
implications for the prudential dimension as well. Specifically, Goldman
claims that:
the account [of causal responsibility] presented here . . . can explain
why people should vote (after obtaining sufficient information) and
it can explain why people do vote (in fairly substantial numbers)
[p. 216].
Taken in context, the idea, as we understand it, is that, once voters reckon
properly the probability of their vote being causally efficacious, they will see
that their vote is likely to have a causal impact on the outcome and will be
able to take some moral satisfaction, after the fact, in having been part of
the cause of their favored candidate’s victory. Furthermore, the prospect
of such moral satisfaction (and/or any social esteem that might attach to
recognition of their role as members of the group that brought about that
candidate’s victory) is a significant incentive.
Interestingly, Goldman and Tuck both talk not just in terms of causal
efficacy but also in terms of causal responsibility. And we shall have some-
thing to say later about the significance of this choice of terms. But we think
the account draws much of its intuitive force from its use of “responsibility”
language, since, with them, we do think that in elections all those who
voted for the victor—not just one pivotal voter—are responsible for the
outcome. So, consider, for example, an individual who voted for Hitler in
the German elections of July 1932. This individual, one might think, and
all the others who voted with her, must bear some responsibility for Hitler’s
successes in that election. Any such voter should properly be held in some
contempt; and she ought to feel guilty for having voted in this way.
Suppose, in her defense, she were to point out that her own individual
vote almost surely made no difference to the electoral outcome. Hitler, she
observes, would have been every bit as successful whether she had voted for
him or not. Would we be—should we be—moved by this observation?
Almost certainly not! Horribly bad moral consequences are associated with
her actions and she ought to take moral responsibility for them.13 And she
ought to do so specifically despite the fact that her individual vote almost
surely made no difference to the electoral outcome.
13
We leave to one side the important question of just how to think of the relative degree
to which she, as compared to others, is responsible for those consequences. We also leave to
one side questions concerning the responsibility she bears for her vote, regardless of the
consequences associated with it. (Here, we are being careful to talk of the consequences
associated with her vote, rather than of the consequences of her vote, since the latter assumes
her vote had the consequences in questions—that it was causally efficacious, and it is that
assumption that we are exploring.)
OUP CORRECTED PROOF – FINAL, 26/03/15, SPi
As will become clear below, we think Goldman and Tuck are right that
there is an important sense in which in elections a group of voters, not just
one, is causally responsible for the outcome. Our concern is with how
thinking about what one might cause figures in justifying and explaining
how people vote. In particular, we want to focus on two questions:
1. What exactly is each J-voter responsible for?
2. And does such responsibility that arises by virtue of being causally
efficacious (or causally responsible) exhaust electoral responsibility
more generally? For example, what account of voter responsibility can
RCV, as we have conceived it, put in play against the Goldman/
Tuck account—and does the best RCV account serve to satisfy our
intuitions in cases like the Hitler-success case?
We shall address those questions in turn in what follows. The upshots of
our discussion are twofold. The first is that there are serious reasons to
doubt whether the Goldman/Tuck account of why people should vote suc-
ceeds in giving the right action-guiding advice. In that sense, the normative
strand of the Goldman/Tuck account is suspect. Our second claim is that
the explanatory version of the account depends on assumptions that the
circumstances of voting make rather dubious; and that under plausible con-
ditions, the effect of the “Goldman/Tuck account of causal responsibility”
on anyone convinced by it may well lead them to be less likely rather than
more likely to vote. In that sense, the explanatory strand of the Goldman/
Tuck argument is suspect.
Both RCV and the rival Goldman/Tuck account are about choices:
the choices individuals do make; and the choices they ought to make. In the
choice context, the proper weight of various considerations bears; and, in
this context, we think two questions loom: “What value is to be placed on
being the cause of some outcome?” and “What probability is to be assigned
to that value?”
The stakes for RCV orthodoxy are considerable. Given that g can be
close to 100% for close elections (and far from negligible for even not so
close ones):
1. We should expect, contra RCV, that the high probability that one’s
vote will be causally efficacious will motivate many people to expend
time and energy in voting. This is, of course, precisely the point that
Goldman and Tuck seek to make.
2. We should correspondingly expect voters to expend significant
effort in acquiring relevant information about candidates and
their policies, contra Downs and Caplan. (Unless, of course, they,
perhaps mistakenly, think that they already have the relevant
information.)
OUP CORRECTED PROOF – FINAL, 26/03/15, SPi
Goldman and Tuck tie their account of moral responsibility to the notion
of causal efficacy. We think that this elision is misleading. We think a proper
account of moral responsibility is often usefully kept apart from any particu-
lar account of causation, and, indeed, in many cases, apart from causation
altogether; and that this is true in the electoral case. Several considerations
incline us to this view.
For one thing, we think that the Goldman/Tuck formulation, by
grounding the idea of moral responsibility in its causal effects on the elec-
toral outcome, fixes the idea of moral responsibility in electoral behavior
too narrowly. It is not that we think that bringing about an electoral out-
come is totally irrelevant to the ethics of voting behavior, but as we shall
argue, it is not the only relevant aspect. More generally, though both respon-
sibility and causality are notions whose application turns on context and
convention, the relevant contexts and conventions are significantly different
in the two cases.
Take causation. It seems clear that in most (perhaps almost all) possible
descriptions of what caused some ‘outcome’ X, there are a number of factors
that might be isolated all of which were present—and necessarily—for X to
14
We believe it is likely that many people do think that as long as they voted for the
winner of an election, they helped to cause the victory, as Goldman’s (but not Tuck’s)
version of g would have it. Yet, we suspect that thought does not depend on ideas con-
cerning their being in a minimally sufficient set. In any case, it does seem plausible that
many people recognize that they cannot claim credit for a victory without voting and
suppose that as long as they do vote (for the winner) they can claim credit. And we think
that is likely a real incentive to many people.
OUP CORRECTED PROOF – FINAL, 26/03/15, SPi
15
Whether she is a part of the cause depends, of course, on whether J wins and also,
on the Goldman/Tuck account, on whether her vote was part of (on Tuck’s version) the
minimal set of votes that was sufficient for J’s victory or (on Goldman’s version) a mini-
mal set of votes sufficient for J’s victory.
16
As Tuck notes acknowledging a concern with agency, and with being oneself the
cause of an outcome, is perfectly compatible with a holding that RCV offers the right
account of when a prospective outcome provides instrumental reason to act. See pp. 54–7.
OUP CORRECTED PROOF – FINAL, 26/03/15, SPi
can be desired for its own sake, we can see that changes might be made even
when the agent does not know or much care whether such changes would be
desirable were they to come about in some other way. A person may think:
“these are my changes and I am making my mark upon the world.” And that
exercise of agency may well be the source of her satisfaction.
Note that, since on the Goldman/Tuck account the reason to vote relates
to being causally responsible for the outcome, the desire to be a cause pro-
vides individuals with incentives to vote for the candidate most likely to
win. This is so because on the Goldman/Tuck account, you cannot be caus-
ally responsible for something that doesn’t happen: voting for a losing can-
didate deprives you of anything to be responsible for.17 So, if you want to be
causally responsible for an event in political history, just for the sake of
leaving your mark, you need to vote for the winner.18
It is worth noting too in passing that the notion of agency at stake here is
distinct from that implicit in most rational actor theory, because in RCT
any and all rational action is seen as the outcome of choice. So any choice is
an example of the individual exercising “agency.” In particular, the choice to
abstain rather than to vote is a choice—and hence abstaining becomes an
action, an instance of agency. By contrast, the notion of “agency” we see as
being brought to the fore in the Goldman/Tuck account of voting has a
more common-sense quality: to play tennis or go jogging is to exercise
agency in a way that choosing to do nothing would not be; to choose to
intervene in the world is to exercise agency where to choose to let things
take their course would not be. And specifically, to vote is to be active in a
process, in a manner that abstaining would be passive. Both voting and
abstaining involve choices; but only one of those choices involves one’s full
political agency. Or at least, so the thought goes. And it is by no means an
unfamiliar thought. The distinction between killing and letting die depends
precisely on a notion of agency in this “common sense” sense. The decision
as to what to do in familiar trolley problems involves a choice; but “agency”
is involved only when you pull levers or push fat men.
Of course, in the CEO/dean cases, and many others, the agent is fully
causally efficacious in bringing about the outcomes that represent her mark
upon the world. That is not so in the voting case. Nevertheless, there can be
17
“On my account,” Tuck writes, “it is rational to vote (all other things being equal)
only if I believe that there are likely to be enough votes for my candidate for my vote to
be part of a causally efficacious set . . . In other words, I think that it is precisely in the
situation where it looks on the standard modern view [RCV] as if my vote is unnecessary
that I have a good reason to vote” (p. 60).
18
For an account of voting behavior based on the related, but importantly different,
conjecture that voters want to be on the winning side for its own sake (at least over some
range), see Schuessler (2000).
OUP CORRECTED PROOF – FINAL, 26/03/15, SPi
Of course, there is some difference in the relative values of J’s victory and of E that
20
would recommend staying and voting. But for a broad range, the fact that J’s victory is
more valuable will in effect be irrelevant to what A has reason to do, given the expected
votes of others.
OUP CORRECTED PROOF – FINAL, 26/03/15, SPi
is expected to vote for J): A’s vote will almost surely not make a difference to
the outcome. That is, h is very small. At the same time, g is substantial and
the Goldman/Tuck emphasis on whether one will be a cause (in their sense)
of a valuable outcome misdirects attention to the possibility of being a cause
of that outcome, ignoring the crucial fact that one is close to decisive over E.
Absent there being an independent (and substantial) value to being a cause
(again, in the Goldman/Tuck sense) of some outcome, a concern for the
value of outcomes ought to lead one to be both indifferent to g and attentive
to h in determining what one should do.
Goldman and Tuck may dispute that A can know with total certainty
that J will win. We, of course, concede that point. Suppose we allow that
A’s absence does reduce the probability that J will win by a certain frac-
tion. So the expected cost of A going out of town is not zero after all.
That does not bring the Goldman/Tuck account of causal responsibility
in line with reasonable action guiding advice, unless the probability that
J might lose is extremely high (given the assumed values at stake). And to
insist that it matters that the probability is not zero is, in effect, to con-
cede what is at stake—namely that what is relevant for determining what
A should do is not the share of so-called “causal responsibility” (as
Goldman/Tuck would have it)—not the probability that A, if she votes,
would have been one of the people that counted to make up the minimal
majority—but rather the likelihood that J will lose because A doesn’t vote.
In short, h.21
The example could be set out, not in terms of voting, but in other terms
where the same issues are in play. Suppose the Coast Guard has two boats—a
small one and a large one. The small one requires only a one-person crew;
the large one requires a six-person crew. There are seven people on call at the
station and two distress calls come in from different points along the coast.
One involves a ship with 140 people; the other involves a boat that has just
one person. All seven could attend to the ship with 140. Or six could crew
the larger boat, while the seventh takes the smaller boat out to save one
more. The best outcome is achieved, we assume, if both boats go out and
141 are saved. We take it that under these circumstances the right action
guiding advice is that someone take the small boat out.
But it is far from clear that this outcome is what the Goldman/Tuck
approach would recommend. Consider the calculus of each of the coast
21
Tuck acknowledges that “it might obviously be a good reason for doing something
that I am probably the only person able to do it” (p. 59), so it may be that our disagree-
ment is more a matter of emphasis, than principle. But we are struck by the extent to
which Tuck (and Goldman) think the relevant consideration in contexts of voting is
whether one will likely be part of a relevant set of votes that is minimally sufficient for
victory, and not whether one’s vote is likely to be necessary.
OUP CORRECTED PROOF – FINAL, 26/03/15, SPi
guards. Each can go with the larger ship and take moral credit for saving 140.
(Just how much credit redounds to each is unclear. One reasonable sugges-
tion is that the credit should be divided equally among all who participate
in the rescue, or among the six in the minimally sufficient set; alternatively,
though, it might be (as Tuck argues) that each rescuer in the minimally
sufficient set can take full credit for saving all 140,22 still another is that the
credit to be shared equally is determined by the size of the minimally suffi-
cient crew.) Or one could crew the smaller rescue-vessel (receiving full
moral credit for saving just one person) and allow her colleagues to crew the
larger ship.
Suppose for the example that, out of gratitude, each person saved pro-
vides a reward of $N to the crew that saves her, and let’s tentatively use that
reward as a proxy either for the moral credit, or for the value, of saving
them. Then the crew of the larger boat will collectively receive $140N (to
be shared in some way by those who are causally efficacious), whereas crew-
ing the smaller boat promises only $N. Absent side payments—and irre-
spective of whether each in the larger group divides the money, credit, or
utility, evenly, or each can claim the total amount of what they cause (as
understood by Goldman and Tuck)—each reasonably prefers to crew the
larger boat. And this remains the case whether the reward comes as money,
or social esteem, or moral satisfaction.23 Of course, there is a reward/credit-
sharing scheme that will create incentives to produce the best outcome
(which, in this case, we take to involve maximizing lives saved). That scheme
will require that the total amount of reward to be distributed across all seven
in a way that is insensitive to who saves whom. This will secure the best
outcome because, when the seventh goes with the larger boat rather than
the smaller, her participation serves to reduce pari passu the reward each of
the six would otherwise receive. The total reward (monetary, social, or
moral) across all who crew would be maximized when all 141 are saved.
And one can certainly imagine a “reward-sharing” scheme that would secure
this result. But note that that “best-outcome” reward-sharing scheme is more
difficult to imagine operating in the case where the reward comes in the form
of esteem or moral satisfaction. Monetary rewards are readily transferable
22
Tuck focuses on the allocation of utility, not moral credit, arguing that in an elec-
tion that has an outcome with a utility of 200, each vote that is part of the causally suffi-
cient set, “represents” a utility of 200. But he cautions that we shouldn’t conclude “that it
possesses such utility, precisely because the notion of representation implies that the
thing represented is different from its representative” (pp. 42–3).
23
The difference between h and g makes all the difference here, since, however, the
credit or utility is to be apportioned, the much higher value of g, compared with h, makes
the expected utility of joining the crew of the larger boat much greater than the value of
crewing the smaller boat.
OUP CORRECTED PROOF – FINAL, 26/03/15, SPi
between persons: esteem and moral satisfaction are not. That is, it seems
plausible to suppose that public esteem and even “moral self-satisfaction”
associated with worthy acts (like saving people in distress) accrues to those
actors who are “causally responsible” whether in whole or in part. Securing
the “incentive-compatible” reward-sharing would require a separation of
reward from agency that seems entirely alien to the spirit of the Goldman/
Tuck treatment.24 The divergence between what action serves to secure the
best outcome and what action serves (on the Goldman/Tuck account) to
maximize the individual actor’s moral credit is troubling. It is troubling not
least because, for Goldman and Tuck, the analysis is supposed to explain
why individuals will have incentives to behave in particular ways—and if
moral credit is indeed allocated according to the Goldman/Tuck scheme,
then it is clear that the outcomes secured will often not be best.
There are really two issues here, as we see it. One concerns which scheme
would provide incentives that would predictably provide better outcomes.
The other concerns which way of thinking about what one might cause by
acting provides the right account of what one has reason to do. (We here
take no stand on how tightly connected these two concerns are, though we
think they are connected.) Our main point is that the weight the Goldman/
Tuck account gives to being (in their sense) a cause of an outcome gives the
wrong action guiding advice, if what matters is the moral (or other) value of
the outcome.
We recognize that these examples may not be decisive. There may be
other examples where the Goldman/Tuck approach might produce better
outcomes than the RCV equivalent. Furthermore, we concede that quality
of outcomes qua outcomes may not exhaust the moral domain. Nevertheless,
it seems to us that failure to provide the right action-guiding advice in
the case we describe (and the many others that have the same structure) is a
significant count against the Goldman/Tuck account. To meet this charge
defenders of the Goldman/Tuck account need arguments for thinking
either that each person should indeed join the larger boat or that their
account does not make that recommendation. As far as we can see, Goldman
and Tuck provide no such arguments. Nor do we see how those might go.
24
One might seek to block the force of the Coast Guard example in a number of ways.
One might be to insist that the seventh is likely not a cause because she adds nothing to
whether the project of saving will be successful, assuming the other six are in place. This
is a point defenders of RCV would make (even as they allow that there is a small chance
that the seventh would prove necessary to success since, for instance, one of the others
might become sick or fall over-board). But clearly, since only six are needed, and so only
six are in the minimally sufficient set, g, for each, will still be extremely high on Tuck’s
interpretation (and on Goldman’s, since each of the seven is a member of some set of
sailors that is sufficient).
OUP CORRECTED PROOF – FINAL, 26/03/15, SPi
Nothing that we have said so far denies that when A votes for J, A is in fact
supporting J’s victory, nor is it to deny that A might have a high probability
of being (in the Goldman/Tuck sense) a cause of J’s victory. What we deny
is that that probability (g) provides much by way of a reason to vote. But
that doesn’t mean that J has no substantial reason to vote. Indeed, we think
people find, and are right to find, significant value in exercising their agency,
in no small part as a way of expressing their political commitments. And we
think people are rightly seen as responsible for exercising their agency in this
way (as in others). Yet, the importance of exercising agency is in play regard-
less of whether one votes for the victor or not. What matters in elections, we
are thinking, is both the outcome and the exercise of agency, but not that in
exercising that agency one happens to be among those who (in the Goldman/
Tuck sense) cause the outcome.
When A votes for J, she does three things: first, she reveals certain things
about her attitudes and beliefs; second, she reveals her preparedness to express
those attitudes and beliefs; and third, she expresses those beliefs specifically
at the ballot box. All of these attributes/actions are ones for which A can
properly be held responsible (and are proper grounds for pride or guilt). The
moral responsibility for those attitudes and expressions applies, over and
above any causal influence on J’s election and, indeed, independently of
whether J wins.
Consider attitudes first. People can be, and typically are, held responsible
for their political attitudes, whether or not those attitudes actually bring
about the states of affairs that are the content of those attitudes. Someone
who holds the view that Hitler’s eugenics policies were admirable surely
does deserve our contempt on that account, even if she never does anything
in relation to those policies other than to admire them. If A’s moral judg-
ments are defective then A is morally defective to that extent.
But expression of such attitudes involves a further step. Sometimes, when
people have odious views, it is best that those people keep such views to them-
selves. They ought to be silent at dinner parties when certain topics come up;
they ought to refrain from writing op-ed pages; or calling in to live radio
shows. It is one thing to hold the views in question, another to express them. If
A gives expression to the attitudes in question—if he declares them—he is
endorsing them, giving them a public life that they do not deserve.
We are drawn to the Goldman/Tuck idea (as we interpret it) that when A
expresses his attitudes at the ballot box he is expressing them in a distinctive
way—in a way that may have causal upshots, however small. For given the
OUP CORRECTED PROOF – FINAL, 26/03/15, SPi
25
And we are concerned that taking g, rather than h, as the probability relevant in
making choices obscures, rather than clarifies, what is at stake. When it is important to
an agent to exercise her agency, as we recognize it might well be, we think it appropriate
to use h, not g, in thinking about the expected value of the options (in these cases not just
the agency-independent value of the potential outcome of that exercise, but also the value
of exercising agency). Of course, choosing to exercise one’s agency is, normally, both
necessary and sufficient for doing so, so h, in such cases, will be 1.
OUP CORRECTED PROOF – FINAL, 26/03/15, SPi
Tuck view most of its intuitive force. Such an account suggests that a voters’
causal efficacy in relation to electoral outcomes is rather less important than
Goldman and Tuck seem to imply. Certainly, on the account of responsi-
bility we’ve highlighted, individuals will systematically be morally responsible
for their beliefs and attitudes, and their expressions of them, and how they
cast their votes, and will remain so whether they vote for the winner or the
loser. At the same time, we suspect that, if one has bent one’s agency in
support of an outcome, and it comes about, one is to some degree respon-
sible for that outcome—whether or not one was (in the Goldman/Tuck
sense) causally efficacious in bringing it about.
The Goldman/Tuck account purports to show not only why agents ought
(or at least have moral or prudential reason) to vote; but also to explain why
significant numbers of them actually do. We have already indicated why the
normative aspect of the argument is questionable. But as we indicated at the
close of section 2, the explanatory and the normative aspects of the account
are less closely linked than Goldman and Tuck suppose. And, indeed, as we
indicated there, the desire to exercise agency in and of itself might well
provide a plausible motive for voting. So, insofar as that desire is what
Goldman and Tuck have in mind by “causal responsibility,” the explanatory
part of the argument seems to proceed intact.
To be sure, that explanatory story requires certain assumptions about the
role of moral factors in agent motivation—the idea specifically that agents
are motivated in part by the prospect of moral satisfaction and social esteem.
These motivational assumptions might be controversial in some RCV cir-
cles but they are ones we broadly endorse, and so we shall not discuss them
further here.
Nevertheless, to the extent that these motivational matters are in play,
it is not entirely clear that they do indeed give rise to higher turnout
than would arise if agents based their voting decisions on the standard
RCV calculus—or some other calculus in which perceived causal efficacy
in producing the electoral outcome to be lower. To give a sense of our
misgivings here, return to the example of agency effects involving the
new dean/CEO, exercising her agency largely for its own sake. She may
well receive satisfaction from “leaving her mark” and she may do so
somewhat independently of whether the mark so left is actually desira-
ble or not. Perhaps she thinks her innovations probably are consequen-
tially desirable; but it may well be that her epistemic warrant for this
belief is somewhat tenuous. But now she reads Goldman and/or Tuck and
OUP CORRECTED PROOF – FINAL, 26/03/15, SPi
realizes that she is more accountable for her actions than she had previously
realized. She now sees that the moral and reputational stakes are higher than
she had thought: her “sense of responsibility” increases.
Accordingly, if there is a serious possibility that the changes she intended
to make may be changes for the worse, then she may well be inhibited in
making them. In the same way, if there is a serious chance that J will turn
out to be an inferior candidate, any increase in perceived responsibility may
well inhibit A from voting. The Goldman/Tuck examples take it as given
that J’s election is independently morally desirable.26 But that assumption
occludes real-life uncertainty about the moral qualities of candidates and
their policies.
Recall that by hypothesis there is a significant vote for K as well as J. Of
course, motives for voting can be various; but in the Goldman/Tuck spirit,
it seems reasonable to think that many of those K voters actually believe
that K, not J, is the morally superior candidate. In the absence of any con-
trary argument, this fact should give the J-voter pause. Confidence in the
proposition that J is indeed superior ought to be somewhat shaken.
And in the face of increased uncertainty about the moral qualities of
options, A’s voting calculus takes on a different hue. After all, we take it that
it is a worse thing to bend your agency to the cause of the worse candidate
than it is to abstain. So the effect of increased uncertainty about whether J
is indeed the superior candidate to K seems bound to reduce turnout. And
the effect of increased responsibility, in the face of a given level of such
uncertainty, seems likely to have a similar effect. Greater probability that
one will cause an outcome—a greater chance of being causally responsible
for the outcome—will encourage individuals to acquire more information
about candidates given that they intend to vote. But whether, given the cost
of information acquisition and the risk of making an error, the greater prob-
ability of responsibility will also induce more individuals to vote seems at
best an open question.27
It is worth noting that increased perceived probability of responsibility
may lead A to abstain even when A is certain that J is the superior candi-
date. If what morally commends J is not that J is good but rather that
26
Goldman considers cases in which someone’s preferred candidate is objectively
worse and argues that when that is true the person does not have (an objective) reason
to vote. (He notes that she may still have a subjective reason, depending on the evi-
dence she has concerning which candidate is better.) But he does not consider, as we
do in what follows, the impact the recognition of the possibility might have on voter
behavior, in light of the more expansive account of causal responsibility he and Tuck
defend (pp. 209–10).
27
What is at stake here is how the moral payoffs to the three prospects—voting for the
better candidate, abstaining, and voting for the worse candidate—each respond to
changes in the perceived likelihood of responsibility.
OUP CORRECTED PROOF – FINAL, 26/03/15, SPi
K is even worse, A may have to wear a certain social and moral oppro-
brium just for voting for a bad candidate. In such a case, increased per-
ceived probability of responsibility for the outcome may leave A wanting
to keep her hands clean—abstention may here too emerge as the pre-
ferred option.28
What greater perceived responsibility will do is to increase the propor-
tion of voters who are morally confident or who are ignorant as to how
ignorant they actually are. Voting is left not only to those who properly see
themselves to be relatively well informed, but also to the opinionated, the
self-deceptive, and those who vote for non-morally grounded reasons.
A heightened sense of responsibility, as Goldman and Tuck think a more
proper view of causal efficacy would generate, will not necessarily lead to
higher turnout. But it probably will lead to a voting body that is more mor-
ally confident, whether because voters have gathered more information or
because of independent psychological factors (whose normative status
seems rather more dubious).
6. CONCLUSION
28
We are grateful to Emma Johnson for this observation.
OUP CORRECTED PROOF – FINAL, 26/03/15, SPi
To the extent that the issue does revolve around agency effects so under-
stood, two further questions arise:
a. Does the inclusion of agency effects give rise to the correct ‘action-
guiding’ advice in electoral contexts?
b. Is it plausible that including such effects will encourage increased
turnout in the circumstances of normal elections?
Our answers to these two questions are negative.
Perhaps a negative response in relation to the first should occasion no
surprise. Agency effects invoke factors that seem more at home in a ‘virtue’
account of electoral behavior than a standard consequentialist one; and it
ought to be expected that adding normative requirements of this voter-
virtue kind might ‘cost’ something in terms of the quality of outcomes qua
outcomes. To be sure, Goldman and Tuck do not cast their “causal respon-
sibility” story in virtue language; but we suspect that that is where their
account properly belongs.
As to the second, explanatory question, the issue hangs on whether see-
ing oneself as more rather than less likely to be responsible for the outcomes
of one’s voting behavior increases the incentive to vote. Much depends on
the weight of guilt/shame when one votes for the “wrong” candidate vis-à-
vis moral satisfaction/pride when one votes the right one. If the latter is less
weighty than the former (as we think most likely) then the best course, in
the face of uncertainty about which is the better candidate, may well be to
abstain. Goldman and Tuck steer clear of such uncertainty in their examples—
but it is difficult to ignore the prospect of this kind of error in fact, or
indeed to account for the patterns of voting behavior that Goldman and
Tuck assume in their examples (and especially for anything other than land-
slide victories) without risk of voter error being part of the story. Of course,
having voters abstain when they would otherwise have voted for the worse
candidate is presumably a good thing. But the expected value of that effect
is properly measured by relying on the parameter that RCV theorists have
long insisted is the proper parameter—namely, h.
Bibliography
Brennan, G. and L. Lomasky (1993). Democracy and Decision, New York: Cambridge
University Press.
Brennan, G. and P. Pettit (1990). “Unveiling the Vote,” British Journal of Political
Science, 20: 311–33.
Caplan, B. (2007). The Myth of the Rational Voter, Princeton: Princeton University
Press.
OUP CORRECTED PROOF – FINAL, 26/03/15, SPi
PA RT I I
POLITICAL LIBERALISM
AND PUBLIC REASON
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
3
Political Liberalism
Its Motivations and Goals
Charles Larmore
1. ORIGINS
I begin by explaining why liberalism in general, by its nature and not merely
by historical accident, is a latecomer among the forms of political life.2 In the
past, political association, seeking as it must always do to unite by means of
binding rules a group of people for the achievement of collective goals, first
took the form of communities organized around some single, authoritative,
more or less regimented definition of the human good. Only if people are
at one in their own ultimate purposes, so it was understandably supposed,
1
Larmore (1987); Rawls (1996).
2
Cf. the discussion of this point in Larmore (1996a), 142–4, 211–13.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
64 Charles Larmore
will there be the social cohesion necessary to achieve the purposes of the
whole. In pre-modern times, the dominant ideas of the ends of existence
were generally religious in character since the world itself was seen more
as a realm of superhuman powers to be propitiated than as an object of
prediction and control. At the same time, those in positions of power were
only too eager to encourage such ideas for the more-than-human author-
ity they could themselves draw from them. Politics was thus religious, and
religion political. To be sure, reigning views of the human good did not
go uncontested. But few people believed that society is possible except on
the basis of some shared and even sacred definition of the meaning of life.
As Lactantius memorably summed up the dogma for the Christian Europe
of late antiquity and the middle ages, “it is the fear of God alone that
secures the mutual society of men, by which life itself is sustained, protected,
and governed.”3
Why this common understanding should have begun to crumble in the
16th and 17th centuries in Europe is a complicated story. No doubt, medie-
val developments—the constant tensions between church and state (sacerdo-
tium and imperium), the succession of reform movements from the 10th
century on that sought to recover the true Christian teaching and gave
increasing importance to inner faith and conscience—played a large part.
Whatever the full explanation, early modern times saw the ever more wide-
spread realization that people reasoning sincerely and carefully about ques-
tions of faith and about the nature of the human good in general are likely
to disagree, often because they differ about what it means in these cases to
reason well. To the long-standing problems of political rule—how to tame
the passions and settle conflicts among interests in order to secure the condi-
tions of social cooperation—was added the need to rethink the basis of polit-
ical authority itself, given the breadth of reasonable disagreement about the
ultimate ends of life. This problem is not one simply of people disagreeing
about religious and ethical questions and about their implications for the
organization of society. That too is an age-old difficulty. At issue was the
widening recognition that reasonable people, by their very exercise of reason,
tend to disagree about these matters. For therein lay a fundamental chal-
lenge to the cogency of any attempt to justify, as all political regimes must
seek to do, the system of rules by which people are to be governed.
Early modern times thus saw as well the emergence of various views
about how to handle this new problem—views, in particular, about how to
bypass the profound religious and ethical disagreements and find in such
factors as self-interest, a regard for the favorable opinion of others, or a sense
3
Lactantius (c. 240–320 ad), De ira Dei, XII: “Timor igitur Dei solus est, qui custodit
hominum inter se societatem, per quam vita ipsa sustinetur, munitur, gubernatur.”
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
of what is right and fair the basis of the principles by which people can live
together peacefully and fruitfully. The problem and its solution were a pre-
occupation of such seminal figures as Bodin and Montaigne, Hobbes and
Bayle. They constitute as well the dominant concern of the liberal tradition
that begins paradigmatically with Locke (though the thinkers just men-
tioned count among its progenitors) and has undergone further key devel-
opments in Kant and Mill and to the present day. The central place in this
tradition of theories of toleration attests to the formative role played by the
effort to reconceive political society in the light of reasonable disagreement
about the makeup of the human good.
Now the cardinal principles of a liberal society—principles such as freedom
of expression and association, equality before the law, rights of political
participation—have to be abstract, since they forego appeal to substantive,
culturally specific notions of the good. Yet, if there is to be the mutual trust
required for people not merely to regard such principles as correct, but to
accept the vulnerability, the dependence on the compliance of others and
the subjection to state enforcement, that comes with living in a society
governed by them, they must still understand themselves as indeed “a peo-
ple,” distinct from other peoples and bound together by an antecedent sense
of community independent of their acknowledgment of those principles.4
This common life can no longer revolve around some definition of the
human good. But it can involve such circumstantial factors as geography
and language, and it will certainly and most importantly include a shared
historical experience, centered on the memory of past conflicts, often
bloody, and the hard-won realization that the substantive ideals that once
defined their political existence should now, as abiding objects of reasonable
disagreement, cease to have this authoritative function.
This is why I said at the beginning that liberalism has inherently the
character of a latecomer. It belongs to the very self-understanding of a lib-
eral political order that it has arrived late on the scene. It would be wrong,
of course, to confuse aspiration and reality, to suppose that liberal ideals
have become fully achieved in today’s so-called “liberal democracies” or to
regard “liberal” as simply an historical category designating the practices
Western societies have devised for dealing with reasonable disagreement
about the human good. All the same, liberalism is a political conception
that emerged as a response to one of the defining experiences of modern
times and that has shaped to a considerable extent the world of today. Even
if, as is not true, there were a liberal society without a pre-liberal past, it
4
Liberal thinkers frequently overlook this need for a prior sense of community.
Present-day communitarian thinkers recognize its importance, but wrongly suppose it
can only consist in some shared conception of the good.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
66 Charles Larmore
would still define itself by the need to avoid the lure of trying to organize its
political life around some specific religious or ethical vision. Precisely
because liberal principles rule out such ways of setting the terms of political
association, they allude to the temptation of doing so. Liberalism always
presents itself as a better account of the nature of just rule than those we are
or have been initially inclined to adopt.
This does not mean, however, that all the different forms the propensity
to align political society on controversial ideals of the human good can take
have been clear from the start. On the contrary, liberal thinkers themselves
from John Locke’s time to our own have often presented their political phi-
losophy in terms of an overarching individualist ethic, committed to cultivat-
ing a critical attitude toward inherited forms of belief and cultural traditions,
to thinking for oneself and working out on one’s own the life one will lead,
that is far more disputable than they have supposed.
That liberal thought has taken this path is not surprising. Individualism
has formed a powerful current in our world. It grows out of basic features of
modern society, especially the market institutions of a capitalist economy. It
is also an understandable response to the expanding recognition, ever since
the 16th century wars of religion, that reasonable people—by which I mean
people exercising their general capacities of reason in good faith and to the
best of their abilities—tend naturally to disagree about the essence of the
human good. Classical liberals such as Locke, Kant, and Mill found it there-
fore plausible to conclude in their different ways that the principles of polit-
ical society should be based on this individualist ethic, that they should
abstract from controversial ideals of the good in order to express thereby the
spirit that should shape the whole of our lives. Our allegiance, they believed,
to any substantial view of the good—to any concrete way of life involving
a specific structure of purposes, significances, and activities, such as a life
shaped by certain cultural traditions or devoted to a particular religion—
can be truly valuable only if we understand such forms of life as ones we
choose, or would choose, from a position of critical detachment. This gen-
eral individualist perspective formed the framework in which they presented
their political philosophy. (The idea of “individuality” plays this role explic-
itly in Mill’s essay On Liberty.) Our status as political subjects or citizens
should be independent of whatever specific conceptions of the human good
we espouse because in that way political principles mirror—as Locke, Kant,
and Mill might have said respectively—the fallibilist, autonomous, or experi-
mental attitude that we as persons should maintain at the deepest level of
our self-understanding.
Things have not stayed so simple, however. Individualist ideas have
themselves become an object of reasonable disagreement. From parts of
the Romantic movement to present-day communitarianism, there has
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
5
For a more detailed account of this Romantic theme, see Larmore (1996b),
Chapter 2, as well as Larmore (1996a), 127–34.
6
Constant, De la liberté des Anciens comparée à celle des Modernes (1819): “Que
[l’autorité] se borne à être juste; nous nous chargerons d’être heureux,” in Constant
(1980), 513.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
68 Charles Larmore
The phenomenon itself seems evident enough. How often, for instance,
do we believe we agree with others about some important matter of life only
to discover, as we discuss the subject further, listen more carefully, and
reflect on what has been said, that we are in reality quite far from being of
the same mind! Since, however, it has long been part of our cultural and
philosophical tradition to believe that reason ensures convergence of opin-
ion, reasonable disagreement can appear to be a notion hard to make sense
of. If people disagree about some subject, given the same evidence, then—
so one can feel impelled to think—they cannot all be reasoning correctly.
How, therefore, can their disagreement count as reasonable? Once they
learn of their disagreement, must they not each, if convinced they are fully
reasonable themselves, reckon that the others are not as reasonable as they?
Or if they believe they are all equally reasonable, must they not each con-
clude that the reasons they supposed they had for their views are inadequate
and that the proper position to adopt, pending new evidence or further
reflection, is to suspend judgment about the matter at hand? At best, rea-
sonable disagreement can thus look like a temporary condition: when peo-
ple who believe they are equally reasonable discover that they disagree, they
would seem obliged to backtrack from their opposing views, and as a result
their disagreement would cease.
A first reply to this objection is that suspending judgment may not be an
option when we are considering, not what to believe about some question
of fact, but which path of life to choose or embrace. Then time and circum-
stance press us to decide one way or the other. We are, moreover, already
embarked on the path that is ours at the moment, and the choice whether
to continue as we have been or to set out in some new direction is one that,
at least implicitly, we cannot escape making.
But in addition and more fundamentally, reasonable disagreement can
be seen as likely and as likely to endure when the subject of discussion, in
whatever domain, is of sufficient complexity and difficulty. When the key
concepts involved can be variously interpreted in plausible but conflicting
ways and when the crucial sorts of considerations that bear on the subject
can be differently weighted, again in plausible but conflicting ways, people
can well come to contrary conclusions if their starting points, that is, the
background of existing beliefs, standards, and interests that appear to them
pertinent, are significantly different. Their disagreement will be reasonable
insofar as they have reasoned properly from their respective points of depar-
ture. Reasonable views are justified views—that is, views one can see reason
to accept, given one’s other convictions—but justified views need not be
true. That is why reasonable disagreement is both reasonable, the different
sides each holding justified positions, and an instance of disagreement, the
different sides presuming that there is a correct answer to the question that
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
70 Charles Larmore
divides them. Here, by the way, I should point out that “seeing a reason,” as
I use the phrase, entails that there really is such a reason, though a reason
one can grasp from within one’s own perspective, and is thus to be distin-
guished from “having a reason,” which may apply even if one is not in a
position to grasp the reason.
To return to the objection I first mentioned: some may think—as does
much of the current epistemological work on reasonable disagreement7—
that when people who consider one another to be equally reasonable (to be
“epistemic peers,” in the language of that work) realize that they disagree on
some matter, despite having before them the same body of evidence, they
then have good reason to question their opinions, suspend judgment, or
revise their conclusions. For they must suppose that someone is making a
mistake, and if they regard one another as equally reasonable, they each
have no grounds for ascribing the mistake to the others rather than to them-
selves. Reasonable disagreement, again it may be thought, ought to induce
self-skepticism, so that the disagreement itself will end. Yet if, as I have been
arguing, being reasonable consists in believing and doing what is justified
given one’s background beliefs, standards, and interests, then another per-
son’s coming reasonably to a different conclusion than we do is not by itself
a reason for us to think that our view may be false. A good reason for us to
doubt our view must be one we can recognize to be such from our own
perspective, and the fact that another person has come to an opposing view
by reasoning well, but from different starting points than ours, is not in
general that sort of reason. We can regard the other person’s view as reason-
able or justified though at the same tine wrong, because of the mistakes we
see in his underlying premises.
The recent epistemological writings to which I referred can make little
sense of the idea of reasonable disagreement because they slight the way the
most notable examples turn on reasonable people having different back-
ground convictions about how to determine what to believe and do. If they
happen to mention such factors, they often stipulate that people who differ
in this regard cannot regard one another as epistemic peers equipped with
the same evidence; as a result, they define the phenomenon out of existence.
Generally, however, they focus on disagreement in perceptual and mathe-
matical judgments, in which (in contrast to judgments about the human
good, for instance) background convictions play little role.
To this defense of the idea of reasonable disagreement one may rejoin
that, in the situations described, we cannot regard the other person’s view as
justified unless we consider his premises to be justified as well and that we
therefore cannot coherently think of his view as justified though wrong.
7
E.g., Christensen (2007).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
8
For more detail, see Larmore (2008), 4–5, 12.
9
Rawls (1996), 56–7.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
72 Charles Larmore
10
On the differences between the two, see Larmore (1996a), Chapter 7, and Larmore
(2008), 141–2.
11
In her critical discussion of my views, Martha Nussbaum (2011) complains that I
slight religious citizens by presenting political liberalism as engaged centrally with
disagreement among reasonable people (32). She therefore quite misunderstands me on
this score, presupposing—as in this context I do not—an opposition between reason and
faith. See also n 27.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
12
There is a lot more to be said about the nature of reason and of reasons than is
relevant from a political point of view. For my broader and indeed metaphysical views on
this subject, see Larmore (2008), Chapter 5, and Larmore (2012b).
13
Rawls (1996), 48–62.
14
Sibley (1953). Rawls refers to Sibley’s essay in Rawls (1996), 49.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
74 Charles Larmore
3. FOUNDATIONS
The impetus for the development of political liberalism has been, I observed,
the growing recognition that the general individualist philosophy on which
classical liberalism relied, as it sought to define a fair political order amid
reasonable disagreement about important religious and ethical questions, is
an object of controversy in its own right. However, it is important to see
that political liberalism is heeding more than simply historical experience in
freeing itself from dependence on an outlook toward life that prizes critical
reflection, autonomy, and experimentation. It is also drawing upon moral
assumptions that imply that this is the proper route to take. After all, why
should not liberal thinkers instead dig in their heels and, observing correctly
15
This failure led, I believe, to various deficiencies in his account of “public reason.”
See Larmore (2012a). See also, more generally, my critique of Rawls in Larmore (2008),
Chapter 6.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
Rawls (1996), 217; see also 137. Here, it is to be noted that “legitimacy” and
17
“justice” are not synonymous, as shown by the fact that laws may be legitimate without
being just and just without being legitimate. The difference between them is that
legitimacy applies to a political regime as a whole and has to do with its exercise of
coercive power being justified—though the two terms are not entirely disjoint since the
conditions under which coercive power may rightly be exercised form part of the domain
of social justice. For more on this point, see Larmore (2013), 293–4.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
76 Charles Larmore
18
Rawls (1996), 50.
19
See, in particular, Larmore (2008), Chapter 6.
20
Such a theory of morality is presented in Scanlon (1998).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
21
Here I follow Max Weber (1972), I.1.§17. For more on this conception of the
political, see Larmore (2013).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
78 Charles Larmore
out for ourselves how we should think and act. People of faith, people
imbued with a sense of tradition, still reflect on the real worth of the reasons
their commitments appear to give them—they often cannot help but do
so—even if their deliberations take place within the bounds of their guiding
assumptions. Nothing in this concept of a person or in the principle of
respect that, as we shall see, builds upon it and grounds the liberal idea of
political legitimacy harbors an allegiance to individualist ideals.
To perceive the connection between these three concepts—person, respect,
and liberal legitimacy—imagine a situation in which we seek to bring about in
others by the threat of force their conformity to a rule of conduct we do not
suppose that they too could see reason from their point of view to impose on
the community. We would, it is true, be appealing to their ability to think and
act for what they take to be reasons, for they cannot be moved by threats
except by seeing reason to fear what we may do if they fail to comply. But our
aim would be only to take advantage of that ability, compelling them thereby
to act in a way that advances the ulterior goals we seek to achieve: the estab-
lishment of social order, the realization of some public good, even the satisfac-
tion of some interest of theirs they cannot properly pursue themselves. We
would not then be valuing their ability to think and act for reasons as some-
thing good in itself. For that would involve treating them in such a way that
they could be moved to abide by the rule, not by the reasons to fear the con-
sequences of non-compliance, but by the very reasons we ourselves have for
imposing the rule. In other words, we would have to appeal to their distinctive
capacity as persons, namely their ability to reflectively evaluate apparent or
purported reasons. If we thus sought to engage their capacity for reflection in
getting them to adhere to some rule of conduct, we would be showing this
capacity of theirs precisely the same intrinsic regard we show our own when we
propose to heed that rule in virtue of concluding that there are good reasons
to enact it. As one might say in echo of Kant, though without the metaphysi-
cal complexities of Kant’s ethics and also (unlike him) with an eye only to
coercive actions, we would then be treating these persons, in their capacity as
rational and reflective beings, no longer solely as means but also as ends.
To respect others as persons in their own right when coercion is involved
is therefore to require that political principles, or at least the fundamental
ones, be as justifiable to them from their perspective as they presumably are
to us.22 We need not suppose the same to hold for those moral principles by
which we evaluate others, even strongly disapprove of how they may act, yet
do so without appealing to the state’s means of coercion; those principles,
we may believe, are morally binding on them, whether or not they can see
22
In the next section I explain the qualification as well as indicate how this initial
formulation itself needs to be modified somewhat.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
23
In arguing that reasonable disagreement is no more of a political problem than
simple disagreement, Joseph Raz (1998) misses this connection between problem and
solution.
24
Rawls himself showed a great reluctance to admit that a principle of respect lies at
the basis of liberal thought, largely because of the many different meanings the idea of
“respect” can have. At the end of Rawls (1971), for instance, he declared that he had not
sought to derive his principles of justice from the notion of respect for persons because
that very notion calls for interpretation, which only a conception of justice can provide
(585–6). The hermeneutic point is well taken. But it does not rule out the possibility that
respect, in a sense to be grasped only in the light of his theory as a whole, is a value on
which that theory substantively depends. We may have to rely on our thinking about
justice or about political legitimacy in order to determine the appropriate sense of
“respect” they embody. Yet, the principle of respect for persons, thus delimited, may still
form the moral foundation of the doctrines themselves.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
80 Charles Larmore
The aim is not to justify liberalism to a radically diverse constituency, one that may
include people who reject liberalism’s most fundamental values. Rather, the aim is
to understand how liberal rights and institutions can be publicly justified to the
constituency of an ideal democratic society.25
Though political liberalism must certainly explain how citizens who hold
liberal values can do so compatibly with disagreeing about so much else, it
is crucial to remember—I repeat a point I have made many times—that
liberalism in general and political liberalism in particular have taken shape
as a response to a problem, namely the political problem posed by “radical
diversity” or by what I have been calling reasonable disagreement.26 Quong
tends to proceed as though the liberal tradition already exists, the question
being whether its guiding ideals and assumptions can be shown to be coher-
ent. But the deeper question is why it exists at all, what problem it has arisen
in order to solve. Without an answer to that question, we cannot under-
stand its underlying motivations, which are what I have been trying to bring
out.27 Still, it is true—on this Quong is right: the solution it offers is not one
that can necessarily be justified to all reasonable people, understood as peo-
ple exercising their rational faculties about moral questions sincerely and to
the best of their abilities. In this, it is unexceptionable since every political
conception excludes, a point to which I return later on.
A second implication is that in a liberal regime the norm of respect does
not have the same sort of standing as the constitutional principles worked
out on its basis. Those principles are legitimate in virtue of being able to be
the object of reasonable agreement and their political authority can accord-
ingly be said to derive from the collective will of the citizens. But the same
is not true of the idea of respect. It must be understood as having a deeper
kind of validity. To be sure, it constitutes a political principle since its
25
Quong (2011), 6.
26
Quong goes so far as to assert that on the internal conception “pluralism [he means
reasonable disagreement, CL] is not a fact about the world which liberal theory must
accommodate. Rather, pluralism is to be understood to be a consequence of liberalism
itself ” (Quong (2011), 139; also 142). Though it is true, as I mentioned earlier, that
liberal ideas of toleration have helped to expand the breadth of reasonable disagreement,
they originally emerged as a response to reasonable disagreement about the nature of the
human good.
27
This is why I also think it a mistake for Martha Nussbaum (2011: 20) to want to
discard from the liberal idea of respect its epistemic component, the reference to what
Rawls called “the burdens of judgment” or I the prevalence of “reasonable disagreement.”
One thereby loses a sense of the problem to which that idea is a response.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
82 Charles Larmore
28
Habermas (1992), 51f, 135, 153.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
84 Charles Larmore
30
Rawls (1996), xlviii, 6, 223.
31
It is also to be expected that different liberal societies, as they come to make respect
for persons their fundamental commitment, will develop as a result of their specific
historical experience somewhat different interpretations of what it entails for the
constitution of a just society.
32
This perspective is to be distinguished from the religious view, very influential in the
historical development of liberal thought, that God’s will is that political society take
shape in accord with human reason.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
Schmitt (1932).
33
Rousseau held that devising a form of association that reconciles political authority
34
with the liberty of each citizen is the “fundamental problem” of politics (Du contrat social,
I.6). This problem is insoluble, contrary to what many liberal thinkers—such as recently
Gerald Gaus (2011: 1–2)—have supposed. The liberal aim is instead to reconcile political
authority with the liberty of those committed to certain fundamental moral principles.
Rousseau himself may have been thinking along such lines, as suggested by his contrast
between natural and moral liberty at the end of Chapter I.8 in the same work.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
86 Charles Larmore
see reason to accept, requires (as indicated in my final statement of the idea)
ensuring that the basic principles of political life be justifiable to such peo-
ple as well, on the assumption—counterfactual in their case—that they too
held this sort of respect to be a fundamental commitment, but given
everything else in their present perspective that they could, compatibly with
that, continue to affirm. Though political liberalism, like every political
conception, inevitably excludes, it also includes the excluded in a qualified
way, and not merely as outsiders to be abided (as other conceptions may
do), but as citizens with a role to play in shaping the terms of association.
In this regard, which is not the least of its appeals, it is, I think, unique
among forms of political exclusion.
I cannot, however, end on so positive a note. There remains a worry,
connected with the argument with which I began this essay, about whether
political liberalism is able to meet the challenges of the contemporary world.
Liberalism in general, I observed, is by its nature and in its self-understand-
ing a latecomer in the history of political conceptions. In explaining that
point, I signaled an important assumption in the liberal project of detach-
ing the rules of political society from substantive religious and ethical prem-
ises and recasting them as a body of abstract principles, first and foremost
that of equal respect for persons as well as those of freedom, equality, and
democratic self-government that derive from it. For that project to succeed,
I argued, the citizens of a liberal society must still regard themselves as form-
ing one people, bound together by a solidarity that is rooted now primarily
in their historical experience, if they are to be willing to expose themselves
to the vulnerabilities of political association. Not by accident, liberalism
developed as a political conception keyed to the modern nation-state.
The question, however, is whether this underlying assumption is still real-
istic and whether the liberal tradition, including political liberalism, has the
resources to deal with the salient problems of our time. In a world dominated
by a rampant globalized capitalism and hurtling, not unconnectedly, toward
environmental disaster, the idea of distinct peoples and sovereign nation-
states appears to be increasingly a thing of the past. States find themselves
more and more the prisoner of worldwide financial markets and multina-
tional corporations. The new digital technology of the media and the Internet,
an integral part of the dynamics of international capital, acts to homogenize
cultural differences and to erase historical memory with the immediacy of
information and entertainment on command. Meanwhile, the air thickens,
the ice-packs melt, the seas rise, and existing nation-states, such as they are,
seem an irrelevance, if not indeed an obstacle, to any way of averting the
impending catastrophe.
Some have thought that the solution to these global problems lies in
globalizing the liberal perspective itself, in working out theories of global
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
Bibliography
Christensen, David (2007). “Epistemology of Disagreement: The Good News,”
Philosophical Review 116, 187–217.
Constant, Benjamin (1980). De la liberté des modernes (Paris: Pluriel).
Gaus, Gerald (2011). The Order of Public Reason (Cambridge: Cambridge University
Press).
Habermas, Jürgen (1992). Faktizität und Geltung (Frankfurt: Suhrkamp).
Larmore, Charles (1987). Patterns of Moral Complexity (Cambridge: Cambridge
University Press).
Larmore, Charles (1996a). The Morals of Modernity (Cambridge: Cambridge
University Press).
Larmore, Charles (1996b). The Romantic Legacy (New York: Columbia University
Press).
Larmore, Charles (2008). The Autonomy of Morality (Cambridge: Cambridge
University Press).
Larmore, Charles (2012a). “Political Liberalism and Public Reason: A Critique of
John Rawls,” in Die Welt der Gründe. XXII. Deutscher Kongress für Philosophie.
Deutsches Jahrbuch für Philosophie, Band IV (Hamburg: Felix Meiner Verlag),
1249–62.
Larmore, Charles (2012b). Vernunft und Subjektivität (Berlin: Suhrkamp).
Larmore, Charles (2013). “What is Political Philosophy?,” Journal of Moral
Philosophy 10, 276–306.
Nussbaum, Martha (2011). “Perfectionist Liberalism and Political Liberalism,”
Philosophy & Public Affairs 39, 3–45.
Quong, Jonathan (2011). Liberalism without Perfection (Oxford: Oxford University
Press).
Rawls, John (1971). A Theory of Justice (Cambridge, MA: Harvard University Press).
Rawls, John (1996). Political Liberalism (New York: Columbia University Press).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
88 Charles Larmore
4
Political Liberalism, Political
Independence, and Moral Authority
Dale Dorsey
As I use the term here, political liberalism holds that an acceptable theory
of political justice cannot be guided by substantive principles of the good
about which reasonable people may disagree. If a particular policy or
institutional action is a requirement of justice, this cannot be because this
policy or action promotes human welfare, well-being, or flourishing.
Morality, on the other hand, isn’t like this. Morality doesn’t bracket sub-
stantive considerations about the value of human life in assigning moral
obligations. It seems perfectly sensible to say that if some action of mine
would render many people’s lives more valuable, this is certainly a moral
reason to do it.
These two facts form the central tension I explore in this chapter. Political
liberalism, or so I claim, comes out with the short end. The general argu-
mentative structure runs like this: first, under the sort of political liberalism
at issue here, individuals can be subject to conflicting political and moral
obligations. At a given time, someone can be morally required to φ and
required as a matter of political justice not to φ. Second, I argue that this
fact creates a problem for political liberalism. Liberals are forced to either
reject a sensible claim about the rational authority of moral requirements,
or adopt a position that is incompatible with one of liberalism’s central
motivations.
90 Dale Dorsey
1
Michael Sandel also uses the “bracketing” terminology in Sandel (1989).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
2
According to Rawls, the fact of reasonable pluralism is “the inevitable outcome of
free human reason.” Rawls, Political Liberalism, 37.
3
Thanks to an anonymous reviewer for highlighting this point.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
92 Dale Dorsey
many different flavors. There are moral reasons, prudential reasons (reasons,
i.e., of self-interest), aesthetic reasons, reasons of etiquette and protocol, etc.
In each of these individual domains, the relevant reasons will help to deter-
mine the deontic status of a particular action from the point of view of that
domain. A moral reason to φ just is the sort of thing that counts in favor of
a moral requirement or justification to φ. This reason could be outweighed,
of course, by a stronger balance of moral reasons not to φ, and so on. With
this conceptual background in mind, the bracketing strategy is a substantive
constraint on the sort of facts that could count, as a matter of political justice
(in contrast, say, to morality), in favor of a particular action, policy, institu-
tion, and so forth. The bracketing strategy will hold that the fact that a
particular policy φ would improve the substantive welfare of the worse-off
is not, or does not provide, a political reason, or reason of justice, to φ. (I
should note here that I’m not offering a theory of what it means for some-
thing to be a reason. I am just stipulating that this is what I mean by “rea-
son.”) Instead, bracketed theories of political justice will hold that political
reasons are instead provided by, among other things, facts of what I’ll call
“non-substantive advantage”: primary goods, capabilities, or resources. In
this way, the bracketing strategy is construed as a constraint on political
theories of justice: any theory of justice that posits, as political reasons, facts
about human flourishing is ruled out.
Moral reasons are like political reasons. A moral reason to φ just is the sort
of thing that counts in favor of the assignment of a particular moral deontic
status to φ: justification, requirement, supererogation, etc. However, unlike
political reasons (assuming political liberalism), moral reasons are occasion-
ally provided by substantive considerations of human welfare, flourishing,
etc. The following view, for instance, seems overwhelmingly plausible:
Welfare as Reason: That φ-ing will improve the quality of life for people (or otherwise
promote their flourishing) is a moral reason to φ, which increases in strength as the
number of people whose quality of life is improved increases, and as the marginal
increase in quality of life increases.
One doesn’t have to be a consequentialist, or even a welfarist, about moral-
ity to accept Welfare as Reason. This reason could be overruled, perhaps even
trivially overruled, by a wide range of agent-centered permissions, restric-
tions, etc. But it seems entirely implausible to believe that our moral obliga-
tions should pay no attention whatsoever to the effect our actions have on
life quality. To see how plausible Welfare as Reason is, consider the following
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
simple example. Imagine you are in a room full of buttons. You are forced
to press some button or other (it could be any button). All other things vis-
à-vis these buttons are equal, save for the fact that one of these buttons
when pressed will substantially increase the life quality or human flourish-
ing of one individual. It would seem wildly implausible to say that one has
no moral reason to press this button. Furthermore, given that one has no
choice but to press some button or other, it would seem highly plausible to
say that this moral reason gives rise to a moral requirement: this reason is
certainly strong enough, and does not appear to be overruled by, any con-
trary moral considerations.
Notice, however, that if we accept the bracketing strategy, the fact that
this particular button increases the flourishing of some individual cannot be
a political reason to press the button. It cannot be, in other words, a reason
in favor of a requirement of justice to do so. This is not to say that there
could be no political reason to do so, especially if to do so would also be to
promote non-substantive advantage. But political reasons, given the brack-
eting strategy, are not constituted by facts about the promotion of welfare or
human flourishing. But, given the relationship between reasons and deontic
requirement (from the perspective of any domain), it would appear that if
moral reasons and political reasons tell in favor of different actions, we
should accept the following principle:
Political Independence: Political requirements may conflict with moral requirements.
One might have a requirement of political justice to φ, and a requirement of
morality not to φ.
Political Independence holds that, in certain cases, political requirements and
moral requirements will conflict. The argument for Political Independence is
relatively straightforward. Certain facts count as moral reasons that do not
also count as political reasons—including facts about the advancement of
substantive well-being or human flourishing. Thus, there are potential cir-
cumstances in which moral reasons will tell in favor of φ-ing, and political
reasons will tell in favor of ψ-ing. In such circumstances one would be
morally required to φ, and required as a matter of political justice to ψ.
Hence, Political Independence.
Actual cases of such divergence do not seem to require any conceptual
back-bending. Imagine, for instance, that a particular individual A stands in
a context in which political reasons operate. Perhaps he is deciding whether
to vote for a particular referendum; perhaps he is deciding on which policy
to support in a city council meeting, etc. Imagine that there are two policy
options, φ and ψ. Imagine that ψ is more strongly supported by the balance
of political reasons: imagine that, with respect to φ-ing, ψ-ing is favored by
the set of considerations that are significant for a bracketed conception of
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
94 Dale Dorsey
political justice. Imagine, however, that φ-ing will, in fact, produce a greater
level of individual well-being for all citizens. This fact, however, is not a
political reason in favor of φ-ing: given the bracketing strategy, political
reasons are not provided by the substantive facts concerning good lives for
citizens. If that’s the case, then it would appear that A has decisive political
reason to ψ, and could (at least in principle) have decisive moral reason to
φ. Hence A’s moral and political obligations would appear, in this case, to
conflict.4 (I consider a method by which to reject Political Independence in
§4. But given the divergent interests of morality and bracketed political
justice, Political Independence is a natural enough conclusion to investigate
further.)
I should note that Political Independence, by itself, should not trouble the
political liberal. After all, we can and do recognize non-moral requirements
that compete with moral requirements in individual cases. Consider, for
instance, prudential requirements. Prudence requires one to maximize one’s
self-interest. But, except on the most implausible theories of individual wel-
fare, advancing one’s own self-interest in a prudentially optimal way will
occasionally require one to perform immoral actions. Such conflict does not
entail any problems whatsoever for a first-order theory of prudence that
generates it. And so the mere fact that political and moral requirements
come apart should not by itself be regarded as a problematic result of the
bracketing strategy. Yet.
The analogy with prudential reasons is helpful and worth spending another
paragraph on. In cases in which prudence commands us to perform some
action and morality commands us to perform some alternative action, it is
natural to ask: “What ought I to do really?” “Should I conform to moral or
prudential obligations?” This is surely an everyday feature of normative
4
One might think that I’m begging the question against political liberalism by insist-
ing that there are distinct facts about the moral point of view that might compete with
political justice. But notice that political liberals are not moral skeptics. For instance, in
a manner to be explored, Barry believes that there is such a thing as moral obligation; it
simply happens that political justice is a part of moral obligation. Furthermore, though
Rawls himself accepts a form of “Kantian Constructivism” about the moral point of
view, he does not deny that there are moral facts. He simply insists that moral facts are
suitably constructed. (See Rawls, “Kantian Constructivism in Moral Theory” in Rawls
(1999).) Hence, insofar as Welfare as Reason is decidedly plausible, any second-order
approach to moral theory should be compatible with it. Furthermore, and perhaps more
importantly, if political liberalism is committed to the denial of Welfare as Reason, this
view is a non-starter.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
inquiry. This question, or set of questions, does not seek to determine either
the moral or prudential importance of performing either action. We already
know, ex hypothesi, their relative moral and prudential importance. Rather,
this question asks about the comparative practical or all-things-considered
authority of prudence and morality: in a case of conflict between prudential
and moral obligations, which has priority?
If Political Independence holds, the same question arises in the case of
political obligations. There are at least some cases in which moral and polit-
ical obligations will conflict. And, hence, it seems perfectly sensible to ask,
in any case of such divergence, whether one ought to conform to moral or
political obligations. But, or so I claim, this question causes very serious
problems for political theories that adopt the bracketing strategy.
Indeed, there are constraints on what might count as an acceptable
answer to this question, especially for the political liberal. First, any such
answer must allow, at the very least, that conforming to moral requirements
is at least all-things-considered justified. Sarah Stroud, for instance, writes:
We are inclined to take the fact that S was morally required to φ to be sufficient all-
things-considered justification for her φing, even if φing was prudentially wrong,
aesthetically unattractive, and so on. We generally accept moral necessity as sufficient
reason for φing in such cases, as we don’t for the deliverances of other evaluative
perspectives. It is as if we took the reason-giving force of moral requirements to be
strong enough to outweigh even the combined force of other evaluative perspectives
on S’s action. (Stroud 1998, 177.)
Stroud here is plumping for a principle one might identify as follows:
Permission: if x is morally required to φ at t, x has sufficient all-things-considered
reason at t to φ.
Imagine how strange it would be to deny Permission.5 Imagine that I am
morally required to perform some action, say, going out of my way to buy
a homeless man a meal. Imagine, however, that there are other countervail-
ing considerations: doing so would render me late to my son’s soccer game.
However, if we assume for the purposes of argument that one is genuinely
morally required to care for this person, it would seem terribly odd to say
that doing so is not at least justified overall. We might say, alternatively, that
other considerations might render other actions, say, showing up to the
soccer game on time, all-things-considered justified also. But it would be
strange to hold that I behaved wrongly in caring for this person given that
5
Full disclosure: some deny Permission. See, for instance, Wolf (1981); Crisp (1996).
However, the plausibility of denying Permission for Wolf and Crisp seems to stem from a
particular vision of the moral point of view, viz., an impartial vision. But, as I try to show
below, political liberalism may be stuck denying Permission for any moral view, so long as
it takes seriously Welfare as Reason—which is non-negotiable.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
96 Dale Dorsey
98 Dale Dorsey
6
The dualist idea is obviously inspired by Sidgwick, who could find no justification
for denying that we always have sufficient reason to conform to moral demands and the
demands of prudence. See Sidgwick (1981).
7
Thanks to an anonymous reviewer.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
8
Reason for doubt: Rawls allows that there are three “levels” of justification of a polit-
ical conception: pro-tanto justification, full justification, and “public justification by
political society.” The first and second levels of justification allow that a political concep-
tion may be justified by a given citizen, but ignored in favor of other nonpolitical values.
However, when conceptions of justice maintain only “full” justification, at best we have
a modus vivendi, not stability for the right reasons. Stability requires “public” justification,
which in turn—as Rawls explicitly notes—requires that political values override nonpo-
litical values. (Rawls 1995, 392.)
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
occasion diverge from political reasons, given that some moral reasons are
culled from an interest in the promotion of human flourishing, is not
enough to guarantee that these moral reasons won’t be overridden by other
morally important concerns. And one such concern might very well be polit-
ical justice itself. For instance, Barry writes: “Justice is, of course, a moral
concept: it is wrong to behave unjustly because that is to breach the terms
of a fair agreement for mutual constraint,” (Barry 1992, 77). Following
Barry, Rawls writes:
While a political conception of justice is, of course, a moral conception, it is a moral
conception worked out for a specific kind of subject, namely, for political, social,
and economic institutions. In particular, justice as fairness is framed to apply to
what I have called the “basic structure” of a modern constitutional democracy.
(Rawls, “Justice as Fairness: Political not Metaphysical” in Rawls (1999), 389.)
Here’s how the rejection of Political Independence, on this view, would work.
Though there are moral reasons to promote human flourishing, in the face
of a contrary requirement of political justice, these moral reasons are over-
ridden by stronger moral reasons, viz., reasons to conform to the require-
ments of political justice. This seems to follow from the suggestion that a
conception of justice is a “moral conception”: if theories of justice are, at
least in part, theories of morality, the requirements of justice, or so it would
appear, just are moral requirements.
If this proposal succeeds, Political Independence fails. But there are impor-
tant questions that must first be answered. How might we conceive of moral
reasons to conform to political justice? First, one might say that the consid-
erations that tell in favor of a political requirement to φ (assuming a brack-
eted theory of political justice) are also considerations that tell in favor of a
moral requirement to φ. This strategy simply adopts a particular account of
the moral point of view: morality cares (in a decisive way) about exactly
those considerations that (bracketed) political justice cares about. Call this
the “non-architectonic” strategy. However, one could accept the “architec-
tonic” strategy. This proposal says that, leaving aside the precise content of
political justice, there is a moral reason—indeed, a decisive moral reason—
to conform to political justice. Morality commands conformity to political
justice simply because that’s what justice requires. If either proposal is correct,
the argument for Political Independence fails, despite the truth of its prem-
ises. I take these proposals in reverse order.
9
Thanks to Steven Wall for a helpful discussion.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
If I believe that counting blades of grass is what makes my life best, and you
act in such a way as to cause me to develop my rational capacities in a way
that does not conform to my conception of the good, though this may
actually increase my welfare, it may also violate a moral reason of respect,
given my own attitudes toward the way in which my life goes best. If the
view on the table is correct, moral obligation is itself bracketed: if a particu-
lar policy affects a set of individuals, this policy cannot be morally required
if it promotes the substantive flourishing of those individuals in ways with
which they disagree. But given that, in a political context, virtually all citi-
zens will be affected, this seems to entail that morality will not require indi-
viduals acting in a political context to promote substantive flourishing given
the fact of reasonable pluralism.
This idea has a sheen of plausibility, but I’m skeptical. Most importantly,
it seems to me that the best explanation for a moral prohibition on impos-
ing your view of welfare on my life is that my own account of what is best
for me is the best sort of evidence of what is actually good for me. We have
a tendency to believe that a person’s welfare cannot be advanced by impos-
ing a state of affairs against which this person is alienated, or which this
person does not endorse. (See, for instance, Railton, “Facts and Values” in
Railton (2003), 47.) But, just for my money, it seems implausible to believe
that morality would entail that we cannot be required to promote someone’s
welfare—no matter how much better that person’s life would go—if that
person simply has a mistaken view about the quality of their lives. Morally
speaking, such a principle sounds repugnant.
Indeed, the failure of this proposal is even more clear when we investigate
the strength reasons of respect must have. To see this in more detail, con-
sider the following case:
Troy: Ten people live in the midst of severe suffering and deprivation. To relieve this
condition, two policies are open. Policy A would alleviate the material deprivation and
suffering for all, but would do nothing more to enhance the quality of their lives. Policy
B, on the other hand, would not simply relieve their material deprivation and suffering,
but would also be to adopt a governmental policy of promoting human flourishing,
viz., the exercise and development of the capacity for human perfection, including
(most importantly) rational capacities (assume this is the right view of human
flourishing, just for the sake of argument). Nine of these people accept and value the
exercise of their rational capacities. A tenth, Troy, rejects the value of exercising his
capacity for perfection, instead preferring to maximize his own experience of pleasure.
Imagine that to adopt Policy B would be to increase the substantive flourishing of all
ten, but would very slightly decrease everyone’s achievement of pleasure and non-
substantive advantage (which is more than made up for, in terms of human flourishing,
by their substantially enhanced rational capacities). Take it as given that only Policy B
would do anything more to improve anyone’s life than Policy A.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
Here, reasons of respect tell very clearly against Policy B. The only reasons
that would motivate Policy B rather than Policy A are those that refer explic-
itly to a particular theory of human flourishing, viz., that to flourish one
should exercise one’s rational capacities. To adopt Policy B violates a reason
of respect: it is to act in such a way as to promote a substantive vision of
Troy’s flourishing or welfare that he rejects. And, hence, if reasons of respect
trump reasons of welfare, Policy A is commanded.
Strictly as a claim about morality, this is simply implausible. To adopt
Policy B would be to improve the lives not just of Troy (despite his own
judgment) but also of nine other people. Given this, however, it seems absurd
to say that Troy’s own false belief about his flourishing would or should
stand in the way of a moral requirement to genuinely improve the lives of all
ten. Imagine that it’s not Troy-versus-nine, but Troy-versus-ninety-nine, or
Troy-versus-999, etc. At some point it begins to look mighty implausible to
say that one is not morally required to act in accordance with reasons of
welfare rather than reasons of respect. Considered simply as a proposal
internal to the moral point of view, holding that reasons of respect trump
reasons of welfare does not pass the smell test.
Troy’s case should convince us of the following: reasons of respect, con-
strued in their most plausible formulation, do not trump reasons of welfare.
There is some point at which reasons of welfare are sufficient to override
reasons not to advance someone’s welfare (i.e., Troy’s) in a way he or she
rejects. And, hence, a crucial supposition upon which the non-architectonic
strategy rests cannot succeed.
I’ll consider three responses here. First, one might think that the sheer
weight of the good lives for the people in question might be enough for
moral reasons of respect to favor Policy B. On this view, reasons of respect
do not rule out actions that promote a particular vision of human flourish-
ing for those who reject this vision so long as sufficient welfare is generated
by the action in question. This may be plausible, but it is no savior for
political liberals who would deny Political Independence. If it’s true, then
reasons of respect cannot always coincide with the demands of bracketed
political justice. It is hard to see how the integrity of any bracketed political
theory could be maintained in light of the suggestion that the promotion of
a substantive vision human welfare is perfectly acceptable as an explicit aim
of political justice so long as enough welfare is promoted by such policies.
Here’s a twist on this suggestion. Perhaps reasons of respect do not count
against a particular policy or action that promotes a substantive vision of
human flourishing for a person that said person rejects if said person
would also assent to the policy despite his or her rejection of this vision of
human flourishing. We might assume that Troy himself is not just a rea-
sonable but a beneficent sort: he’s willing to accept a policy that would
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
promote a particular conception of welfare not his own if it’s the case that
in so doing many others conform to a life they believe is worthwhile. So,
in other words, though he does not endorse the value of his rational capac-
ities per se, he does endorse the plan’s proposal given the attitudes of others.
And hence, or so it may be claimed, there is no reason of respect to refrain
from benefiting Troy in this case.
I think there are two problems with this suggestion. The first is that Troy’s
assent to Policy B does not seem to pass the test of stability: given the case
as imagined Troy does not consent to Policy B for the right reasons, but
instead simply given its instrumental benefits. But leaving this aside, the
biggest problem with this suggestion is that even if in Troy’s case, given Troy’s
dispositions, there is no reason of respect not to advance Policy B, not all
people have Troy’s attitudes. Troy could reasonably take a different line. Troy
could refuse to grant Policy B his support given that he believes that it will
do nothing to advance the flourishing of anyone, would be to grant sub-
stantive governmental support to a conception of the good that is not his
own, would cost him a modicum of pleasure, which he values, and would
reduce his non-substantive advantage. And, hence, even if Troy himself
were to grant support to the policy, this would do nothing to guarantee that
reasons of respect, even if such support were relevant, would plausibly
trump moral reasons in other cases.
Third, and finally, one might hold that reasons of respect—construed as
reasons against the promotion of a substantive vision of human flourishing
for individual agents that they reject—are not the only reasons that may
work to yield a denial of Political Independence. In other words, it may be
that not just reasons of respect, but many other potential reasons as well,
have moral weight, yield moral reasons to act as bracketed political justice
requires, and (when combined) can trump reasons of welfare. But Troy’s
case sheds doubt on any such suggestion. This is because Policy A is very
clearly superior when it comes to those things that bracketed political jus-
tice cares about—including non-substantive advantage—and inferior only
when it comes to those things that bracketed political justice very clearly
doesn’t care about, viz., substantive human flourishing. So, however one
understands the moral reasons that command us to act as bracketed justice
requires, no such account of these reasons will plausibly trump reasons of
welfare, as seen in Troy’s case. Morally speaking, Policy B is required. To say
otherwise is to adopt a distorted vision of the moral point of view simply for
the sake of denying Political Independence, and as such can do nothing to
support the non-architectonic strategy.
In sum, I think Troy’s case sheds substantial doubt on the non-architec-
tonic strategy no matter how one seeks to accommodate it. That cases like
Troy’s exist should come as no surprise. To reject Political Independence, the
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
political liberal must thread an extremely fine needle: she must show that
there are no cases in which an independently plausible moral theory would
require us to act contrary to a bracketed theory of political justice. But to do
this she must not only argue that there are independent reasons to accept
a set of moral reasons that tell exactly in favor of those actions supported
by bracketed political justice—which itself would seem cosmically coinci-
dental, but which I have granted for the sake of argument—and that these
reasons, whatever they are, have the moral power to override reasons of
welfare in every possible case. But this is an extraordinarily strong thesis
independently of a substantive consideration of its plausibility. And hence
that there are counterexamples, potentially a number of counterexamples,
should not be difficult to swallow.
Given the failure both of the architectonic and non-architectonic strate-
gies, it would appear that Political Independence (on the assumption of a
bracketed theory of political justice) is well supported. But this leaves the
political liberal in exactly the pickle I presented at the conclusion of §3:
the political liberal, because she must accept the stability constraint, and can-
not deny Political Independence plausibly, must reject the rationalist constraint.
But to reject the rationalist constraint is to deny not just a very plausible prin-
ciple of the relationship between morality and how we ought to live, it is to
deny a principle that most look upon as little more than platitudinous. I claim,
therefore, that the best option is to reject the stability constraint.
In this chapter I’ve explored three mutually incompatible ideas, viz., Political
Independence, the stability constraint, and the rationalist constraint. A brack-
eted theory of justice is stuck with Political Independence. A denial of the
rationalist constraint is implausible. And, hence, it would seem like the best
option is to deny the stability constraint. However, one might question
whether this is really fatal to political liberalism, as I understand it here.
Couldn’t there be other motivations for a bracketed theory of political justice?
Ultimately, I do not rule out the possibility that political liberalism could
survive the critique I offer here by outlining an alternative rationale and
instead simply abandoning the importance of stability for the right reasons.
I leave this as an officially open possibility. But I think that Political
Independence, combined with the rationalist constraint, yields a very impor-
tant challenge for any bracketed theory of political justice, even if we don’t
accept stability as the primary rationale. If we accept Political Independence
and Permission, this entails that it is perfectly permissible—compatible with
all-things-considered obligations—for people, including those who set the
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
terms of public policy and the rules enforced by state apparatus, to design
political institutions and public policy in accordance with moral obliga-
tions, which refuse to bracket substantive considerations of well-being. But
if lawmakers and designers of the basic institutions of society act in a per-
fectly justified manner in promoting a particular vision of human flourish-
ing, a rationale for political liberalism must offer an explanation of why it
would be impermissible for political agents to do so simply on political
grounds.
But this explanatory burden, or so I claim, is heavy indeed. The weak-
ened normative authority of political justice in fact wrecks a number of
possible rationales for a bracketed theory: Political Independence and
Permission entail that even a bracketed theory of political justice will not
help to satisfy a desire “to live in a society whose members all freely accept
its rules of justice and its major institutions,” (Barry 1992, 164) nor does it
guarantee that political institutions will be compatible with the full exercise
of autonomy (Sher 1997, chs 3–4),10 nor does it guarantee that “citizens will
want to propose and abide by” the rules of justice (Quong 2011, 143), if any
of these ideas are construed in a manner that would support a bracketed
theory of political justice. (Of course, one could construe the importance of
agreement, say, as compatible with political institutions being guided by
moral requirements on occasion—but given the influence human flourish-
ing has over moral requirements, this rationale would obviously not support
political liberalism as I construe it here.)
So while I leave open the possibility that there may be some additional
rationale to support bracketing, any such rationale must be able to preserve
the importance of accepting a bracketed theory of justice in light of the fact
that political actors do not necessarily act wrongly in designing their insti-
tutions in accordance with considerations of substantive human flourish-
ing. In my view, we have reason to be skeptical of this possibility, but I’ll not
press the point further.
6. CONCLUSION
10
Of course, this is controversial insofar as (a) the “autonomy” rationale seems incom-
patible with the bracketing strategy as I understand it, insofar as it treats the promotion
of substantive autonomy as a political reason and (b) Sher argues, plausibly, that even an
appeal to autonomy cannot justify bracketing substantive considerations of the human
good when it comes to public policy.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
so-called political liberals, must either reject this strategy or reject the
rationalist constraint. There is no feasible method by which to retain the
rationalist constraint and also maintain the fundamental motivation for
the bracketing strategy, i.e., the insistence on stability of political principles
for the right reasons.
As I’ve noted throughout, there are ways that political liberals might
respond to the argument I offer here that preserve their position. First, they
might accept a stronger form of moral anti-rationalism, and deny Permission.
Second, they might take up the project of offering an independently plausi-
ble (independent, that is, of whether Political Independence is true or not)
account of the moral point of view with an eye to demonstrating that Political
Independence should be rejected despite Welfare as Reason. Alternatively, and
more radically, one could deny the stability constraint, hold that the stability
of a system of political justice need only be guaranteed perhaps in terms of a
modus vivendi, and instead offer some other motivation for a bracketed the-
ory of political justice. Each of these proposals has significant challenges (the
first a challenge of plausibility, the second a challenge of, as it were, “high
odds,” the third the challenge of outlining a rationale that is unaffected by
the fact that political institutions will be designed by upright people in
accordance with facts of human flourishing even if political justice says they
should not do so). In any event, while there are certainly options available for
the political liberal, I have tried to bring out the essential challenges that this
view faces in light of the competing authority of moral and political demands.
Bibliography
Barry, Brian (1992). Justice as Impartiality. Oxford: Oxford University Press.
Crisp, Roger (1996). “The Dualism of Practical Reason” in Proceedings of the
Aristotelian Society, 96.
Quong, Jonathan (2011). Liberalism without Perfection. Oxford: Oxford University
Press.
Railton, Peter (2003). Facts, Values, and Norms. Cambridge: Cambridge University
Press.
Rawls, John (1971). A Theory of Justice. Cambridge, MA: Harvard University Press.
Rawls, John (1995). Political Liberalism. New York, NY: Columbia University Press.
Rawls, John (1999). Collected Papers, ed. Freeman. Cambridge, MA: Harvard
University Press.
Sandel, Michael (1989). “Moral Argument and Liberal Toleration: Abortion and
Homosexuality” in California Law Review, 77.
Sher, George (1997). Beyond Neutrality. Cambridge: Cambridge University Press.
Sidgwick, Henry (1981) [1907]. The Methods of Ethics. Indianapolis, IN: Hackett
Publishing Company, 7th edn.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
5
Against Public Reason
David Enoch
I don’t know of any other philosophical discussion that is quite like this: talk
to (broadly speaking) Rawlsians, and you are likely to get the impression
that some kind of political liberalism, or a public-reason account, is the only
game in town.* The only questions worth discussing, it seems, are within
this framework, rather than about it. Theorists who reject this framework
are often ignored, and the feeling one gets is that they just don’t get it. Talk
to many others (non-Rawlsian political philosophers, as well as philoso-
phers whose main area is not in political philosophy exactly), and you are
likely to get the impression that Rawlsian public reason has been effectively
refuted several times over and, indeed, that even this much was never nec-
essary, as the theory was a non-starter to begin with. And the feeling one
gets is that Rawlsians just don’t get it, and that their tremendous influence in
political philosophy is corrupting the field.
As my title suggests, I am much closer to the second stance than to the
first (though both have been presented here in rather extreme, somewhat
caricatured, forms). So, in this chapter I make yet another attempt to explain
why public-reason accounts are hopeless. I share the feeling that there are
already in the literature strong reasons to reject specific public-reason
accounts and perhaps that tradition more generally, but I think that more
can be done—and that given the state of the field, more should be done.
In this chapter, then, I try to develop some of the strongest, most general
* Earlier versions of this chapter were presented at the ANU, Aarhus University
Copenhagen, Leeds, Arizona, Yale, CUNY, Rutgers, Tel Aviv, and Jerusalem. I thank the
participants on those occasions for the discussion and their valuable comments. And for
extremely helpful written comments, I would like to thank Dani Attas, Dan Baras, Ilan
Ben Shalom, Steve Darwall, David Estlund, Ruth Gavison, Alon Harel, Shelly Kagan,
Barak Medina, Shmulik Nili, Shai Perry, Jon Quong, Chad van Schoelant, Steve Wall,
Alex Zakaras, and an anonymous reader for Oxford Studies in Political Philosophy.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
1
Two I don’t address here: self-defeat (which I address, following others to an extent,
elsewhere; see Wall (2002), and my (2013)); and a critical discussion of the epistemic
commitments of public-reason theorists, which I intend to address in future work.
2
I do some of this elsewhere, with regard to Estlund (2008; see my (2009a)), Gaus
(2011; see my (2013)), and to an extent Waldron (1999; see my (2007)). (In what follows,
from time to time I borrow a few sentences from these other texts of mine.) Almost all
other critics of this tradition target primarily Rawls, of course.
3
For my purposes here it won’t be necessary to distinguish between public-reason
theorists who (following Rawls) restrict the scope of this requirement to just, roughly
speaking, constitutional essentials, and those who (like Quong) apply the requirement
more broadly.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
that we accept it. This means that the impression sometimes highlighted by
public-reason theorists that by flouting a public justifiability requirement
we unacceptably give our own views extra weight is confused. In section 5,
I take a step back from the details of some of the earlier sections, reflecting
on the role of the political philosopher according to public-reason theorists.
I highlight what I take to be a deep, dangerous mistake—among other
things, this conception of the role of political philosophers divorces them
entirely from even the just struggles of political activists. In the concluding
section I return to the starting point—to the powerful underlying intuition
that public-reason theorists, I think, get right. I show how such intuitions
can be accommodated without endorsing a public-reason account.
Suppose that Catholicism is the true religion, and that the Pope has a direct
line to God. Still, if you are a non-believer (or a believer of some other
religion), we find it deeply objectionable for the Pope’s directives to have
authority over you, simply in virtue of the truth of Catholicism and the Pope’s
excellent epistemic credentials.4 Certainly, the thought that you can be
politically subject to the Pope’s directives—backed up by the state’s coercive
power—seems unacceptable. The reason, it seems, is that while his direc-
tives are in some sense justified, they are not justified to you, non-Catholic
as you are. And in order to reconcile legitimate authority with your nature
as a free, autonomous agent, there must be something to be said for it, and
furthermore, this something must be sayable to you. Truth, it seems, is never
sufficient for legitimacy. Rather, those over whom authority is claimed are
entitled to answers, to justifications of the authority that are available to
them, pretty much as they are.
The general thought seems to be that given the liberty, autonomy, ration-
ality, or some such of those subject to the purported authority, its legitimacy
must be somehow accessible to them, it must be such that they too can
appreciate it, it must be such that they have reason to endorse it, or, as the
slogan goes, it must be justifiable to each and every one of them. It is impos-
sible, the thought seems to be, for a free person to be subject to a legitimate
authority if there’s nothing that can be said to him, in terms he can relate to,
as it were, that justifies subjecting him to that authority. Notice that no
consent is required—perhaps, for instance, it’s not about engaging your will
(as consent presumably is) but your reason (a central distinction in Waldron
4
Estlund’s (2008, 5) example.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
(1987)).5 Still, the natural thought is that unless an authority can be justified
to you pretty much as you are, it does not have legitimacy over you.
For brevity, I am going to put this underlying thought as the attempt to
reconcile authority with liberty. Of course, these terms (“authority” and
“liberty”) are neither univocal nor philosophically transparent.6 But for my
purposes here the somewhat loose characterization above suffices—clearly,
something like the tension between authority and liberty does serious work
here, as otherwise it would be entirely mysterious why we should justify an
authority to those subject to it rather than to others. It is they who are owed
justification, because it is their liberty (or some such) that is at stake when
they are subjected to an authority. In my criticism of public reason below,
nowhere will I rely on a more specific, more controversial characterization
of the notion of liberty7 or of the need to reconcile liberty and authority.
Furthermore—and this brings us to the second underlying public-reason
intuition—for the Pope (or any of his followers) to apply his Catholic-based
directives to you non-believer as you are, will amount to giving more polit-
ical weight to his (or their) beliefs than to yours, and so will amount not
only to failing to treat you as free, but also to failing to treat you as an equal
citizen, one whose beliefs, principles, desires, projects are as politically sig-
nificant as anyone else’s. When we combine these two underlying intui-
tions, we get the slogan—the justifiability-to requirement follows from the
commitment to treat each other as free and equal.
This, then, is what in this chapter I take to be characteristic of public-reason
theories. They are all committed to some kind of accessibility requirement
as a necessary condition for legitimacy: the thought that for a state (for
instance) to be legitimate, its authority must be justifiable to each of those
subject to it. And—and this will be important in what follows—they all
share the underlying motivation of reconciling liberty and equality with
authority. To an extent, you should feel free to treat this characterization as
a stipulation—in what follows, I will use the words “public-reason accounts”
to designate theories that are committed to some justification-to condition,
motivated by the need to reconcile authority with liberty and equality. But
5
In fact, we can think about consent as a particular instance of this more general,
arguably more accurate, condition. For if you’ve consented to an authority, we are guar-
anteed to have something to say to you about its legitimacy: namely, that you’ve given
your consent. According to this line of thought, then, the problem with consent theories
is simply that they over-generalize.
6
For my take on authority in general, see my “Authority and Reason-Giving”
(2014).
7
In particular, I will not be assuming a general freedom-to-do-as-one-chooses con-
ception of liberty.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
I don’t think that this way of using words is at all idiosyncratic, and I think
that this characterization nicely captures pretty much all and only accounts
that are usually thought of as members of this tradition. Here, for instance,
is Nagel (1991, 33–6):8
The task of discovering the conditions of legitimacy is traditionally conceived as
that of finding a way to justify a political system to everyone who is required to live
under it . . . the search for legitimacy can be thought of as an attempt to realise some
of the values of voluntary participation in a system of institutions that is unavoidably
compulsory.
And similar characterizations of the public-reason account and underlying
motivations are common.9
Now, as noted above, my hope is to criticize the entire public-reason
tradition, not any specific members thereof. So I cannot rely on more
specific details—say, the exact nature of the accessibility or justification-to
requirement.10 Though such differences may be important in other con-
texts, for my purposes here what’s important is what such different accounts
have in common. In particular, then, I can afford not to say much more
about what exactly it is that is needed for it to be the case that something is
not just justified, but is justified to the relevant constituency. Still, one thing
must be emphasized here: though I am going to remain neutral on the
precise nature of that extra thing that is needed (for justification-to), I am
going to assume that this requirement is non-vacuous—that is, that a justi-
fication that is sound in an important sense can be accessible (in the relevant
sense) to some but not to others, that the justification-to requirement takes
8
Quoted in a similar context in Rossi (2014).
9
For statements of an accessibility or justification-to requirement as the definitive
feature of this tradition, see, for instance, Waldron (1987, 128); Quong (2011, 161);
Quong (2013, the opening line); Gaus’s (2011, 263) “Basic Principle of Public
Justification;” Vallier and D’Agostino’s (2013) Public Justification Principle; and Wall
(2002, 385, 387).
For a helpful survey of possible motivations for public-reason accounts, see Quong
(2013, section 1), and the references there. There, thoughts about coercion constitute
only one section out of five. But it is, I believe, the central one, and much of what goes
on in the other sections also tacitly relies on it. For instance, another motivation Quong
brings from Gaus has to do with Gaus’s account of our reactive attitudes, and some acces-
sibility being necessary for their appropriateness. But for Gaus, the appropriateness of the
reactive attitudes is closely tied to thoughts about when the relevant part of morality is
objectionably authoritarian—that is, to an extent, inconsistent with the liberty and
autonomy of those subject to it.
10
For a helpful list of relevant distinguishing questions here (the nature of the accessibil-
ity, the constituency, the nature of the thing to be justified, etc.) and for many references,
see Vallier and D’Agostino, section 2.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
as input something about the actual features of the relevant addressee.11 For
now, this will do.12
3. IDEALIZATION
11
Raz (1998) claims that the only accessibility requirement that makes sense here is
one that is always trivially met, because all good normative arguments are always accessi-
ble to all. I don’t agree with Raz on this—I think that there is sense to be made of the
thought that the Pope’s justifications are not accessible to a non-believer, even under the
assumption that Catholicism is true. But my point in the text is just that an account that
superficially uses the justification-to lingo, but which happily takes on board this point
from Raz, is not within the target of my arguments in this chapter (nor is it naturally
classified as a public-reason account).
12
The words “public reason” are sometimes used in a different—narrower, perhaps
even more precise—sense, one having to do with the shared language of officials and the
Rawlsian duty of civility not to rely on private conceptions of justice. See Quong (2011,
41–2), and the references there. And for the distinction between the two ways of using
“public reason,” see Quong (2011, 256). I have nothing here to say about this other sense
of “public reason”—I just note it here in order to preempt confusion.
13
Indeed, arguably the nature of the idealizing device doesn’t much matter for the
points to come. In particular, there is structural similarity between excluding the unrea-
sonable and going hypothetical, perhaps to the point that a more theoretically parsimo-
nious public-reason theory will make do with just one of those (say, packing enough into
the notion of the “reasonable” so that once the constituency has been restricted to the
reasonable, no further hypotheticalization is needed). I discuss these two mechanisms
mostly because public-reason theorists employ both, usually in tandem. It should come
as no surprise that similar worries arise regarding both.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
14
See, for instance, Larmore (2003); Quong (2011).
15
See, for instance, Gaus (2011).
16
Rawls, and following him pretty much everyone else, talks of reasonableness. Talk
of qualified acceptance comes from Estlund (2008).
17
See Gaus (2011, 250).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
18
The line here echoes a well-known criticism of hypothetical consent theory, for
instance, in our context, Dworkin (1973).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
19
See my “Why Idealize” (2005). But see also Sobel (2009) and my (manuscript).
20
Worries in this vicinity have been around since very early on. See, for instance,
Hampton (1993, 299), and Raz (1998, for instance 37). In what follows, I try to address
them both more generally (not just vis-à-vis Rawls), and in a more principled way
(based on the general thoughts in the text about when an idealization is philosophically
acceptable).
21
See, for instance, Quong’s (2011, 291) understanding of (Rawls’s understanding of )
the reasonable.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
22
In this respect, Estlund’s insistence on going for a more technical term (“qualified
acceptability”)—motivated precisely by the kind of considerations in the text here—is
more helpful.
23
For an especially clear example, see Quong (2011, 156).
24
Quong’s examples, for instance, are those making “claims to racial, gender, or ethnic
superiority” (292), the Ku Klux Klan (299), white supremacists (309), Nazis (309), and
psychopaths (314).
25
I address the shaky epistemological-sounding claims made by Rawlsians in work in
progress, provisionally entitled “Political Philosophy And Epistemology: The Case of
Public Reason.”
26
For explicit claims to this effect, see Dreben (2003, 326); Estlund (2008, 61);
Quong (235, fn 34; 240). As far as I know, Rawls nowhere says this explicitly. Nor do
I know of a place where he rejects this claim.
27
For instance, I think—but I’m not entirely sure—that according to Quong’s con-
sensus view, a Gaus-like convergence view is unreasonable.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
such people as John Stuart Mill, Karl Marx, Joseph Raz, Jean Hampton,
pretty much all contemporary epistemologists, probably most of those
offering a rival public-reason account, presumably the early Rawls—oh,
and me.28
Now, a case can be made that these aspects of the delineation of the
unreasonable are a necessary feature of public-reason accounts.29 But, loyal
to my attempt to target the tradition in general rather than some (even all
developed) specific manifestations thereof, let me not commit myself to this
strong claim. So, the thing to keep in mind is not necessarily that the rea-
sonable include all non-Rawlsians, but that they include many more than
just the Nazis and the murderous psychopaths.30
Can public-reason theorists retreat to a pre-theoretical understanding of
the reasonable, perhaps so that you qualify as reasonable if your reasoning
mechanisms are functioning well, or if you’re willing to listen to other views,
and so on?31 They can, of course, but the price would be anarchism again—
for among those who are reasonable in just some very thin sense of this
kind, everything is controversial. Under any plausible hypotheticalization,
and any non-trivial understanding of accessibility, nothing is justifiable to all
the reasonable in this sense.
Keeping in mind, then, how easy it is to be unreasonable, let’s return to
the philosophical motivation underlying public-reason accounts—the need
28
Public reason theorists can insist (as Quong (2011, Chapter 10) explicitly does) that
the unreasonable should enjoy the same rights that the reasonable enjoy—it’s just that
which rights these are is going to be determined by justifications that are accessible only
to the reasonable. Still, even when just talking about the distinction between those to
whom justification is owed and those to whom it is not owed it is important to keep in
mind that the latter include many who are, in the natural-language meaning of the term,
perfectly reasonable.
I should note that Quong’s commitment to equal rights to the unreasonable is not
without qualification. Quong is willing to seriously restrict the freedom of speech of the
unreasonable (309). I am not sure, but I think that Quong may be committed to the
justifiability of preventing me from publishing this article, for instance.
29
At least one plausible way of going on the problem of self-defeat has it come out as
a theorem that all non-public-reason-theorists are unreasonable. But perhaps there are
ways of avoiding such a result.
30
The point is not that there’s something strongly counterintuitive about classifying
these people as unreasonable. The point, rather, is first, that the public-reason theorist
can’t have it both ways—if he applies everyday natural-language intuitions regarding
“reasonable” in ways that support his theory (which he shouldn’t do, as “reasonable” here
is a technical term), he must say something about the cases in which they clearly challenge
it; and second, that in the technical sense of “unreasonable”—namely, those justification
to whom is not a necessary condition of legitimacy—excluding all those people is indeed
counterintuitive.
31
Sometimes, public-reason theorists write as if they do. See, for instance, Nagel
(2003, 76); Larmore (2003, throughout).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
to reconcile authority with liberty and equality. And we must ask—do such
thoughts apply to the unreasonable as well? The answer seems obvious. The
unreasonable too are born free, and yet they are everywhere in chains.32 And
if by coercing someone based on a principle they do not endorse (nor would
they, after a respectable amount of good reasoning) we fail to treat her as our
equal, then this is so when that someone is unreasonable as well. If you’re
not sure, recall the need to resist the temptation to think of the unreasona-
ble as the Nazis and the murderous psychopaths. Suppose I am quite, well,
reasonable (in the non-technical, natural language sense of this term),
except I reject some of the burdens of judgment. Is there no longer even an
initial tension between my liberty (in whatever sense is important here) and
the authority the state claims over me? Suppose I am motivated to interact
with others as free and equal, but I have other stronger motivations, or
perhaps my conception of what it is to treat others as free and equal is a
little bit off. Is there no longer a problem of equality if you impose on me a
principle I do not endorse? Alternatively, just think of some of the smartest,
nicest, people whom public-reason theorists rule out as unreasonable, and
ask whether the underlying thoughts about liberty and equality apply to
them, at least initially. I can’t see how a negative answer can even get off the
ground. A story reconciling authority with liberty and equality is still very
much needed, with regard to (at least many of ) the unreasonable as with
regard to everyone else.
If the initial problem arises with regard to the unreasonable as well, and
if the restriction to the reasonable is not to be an ad hoc restriction intro-
duced just in order to save the theory, some other rationale must be sup-
plied for the exclusion of the unreasonable. For the most part, this is a need
left unacknowledged by public-reason theorists. Still, at times such ration-
ales are offered (or can be read off what public-reason theorists say). Let me
address the two main ones I find in the literature.33
It is sometimes said that reciprocity considerations are what does the trick
here. The thought seems to be that we should justify ourselves only to those
who, had they had political power, would have bothered to justify them-
selves to us.34 Now, it seems to me an interesting, general question how such
32
I’m paraphrasing Rousseau here, without any exegetical pretensions.
33
It is sometimes hinted that practical considerations—having to do, roughly, with
the costs of including the unreasonable—do the work here. See, for instance, Cohen
(2009, 7), Quong (2011, 37). But I don’t think that these suggestions are in line with the
public-reason spirit—on top of relying on questionable empirical assumptions (seeing
how many people are unreasonable in the relevant sense, excluding the unreasonable may
also be extremely costly), it seems too modus-vivendi-ish for Rawlsians. This is not what
they seem to be after.
34
Macedo (1995, 484), Larmore (2003), Lenman (2010, 179).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
priority (in some sense) over non-ideal theory. But even granting Rawls all
that, still we shouldn’t be tempted by too quick a move from ideal to
non-ideal theory. Even if Rawls is right that some kind of justifiability-to
is a necessary condition for legitimacy in an all-reasonable-community,
it most certainly does not follow that in a community where some are
unreasonable, the analogous condition necessary for legitimacy is
justifiability-to-just-the-reasonable (nor does it follow, if you’ve justified
something to all the reasonable under full compliance, that it’s also justi-
fiable—even just to them—under partial compliance). Unless some
rationale can be offered for this restriction in the non-ideal theory, then
perhaps the thing to conclude is that no justifiability-to requirement is
necessary for legitimacy in non-ideal theory, because none is satisfiable
(because of the presence of the unreasonable). Or perhaps there are some
other ways to go. The general point is simply that it’s never a simple,
immediate step from ideal to non-ideal theory.36 Ignoring this would be
tantamount to trying to design airplanes by relying on mechanical mod-
els that assume frictionless surfaces, just because there’s something to be
said for thinking of some such models for some purposes. So, despite the
intellectual respectability of the ideal-theory way of motivating conduct-
ing a discussion that ignores the unreasonable, this way of proceeding
cannot at the end of the day vindicate excluding the unreasonable, at least
when we’re back to talking (as public-reason theorists seem to) about the
real world.37
36
For similar points in similar contexts, See Cuneo (2013, 360), and Rossi (2014).
This seems to be a point Quong understands well, at least when criticizing others:
“ . . . even if we stipulate that people can only be members of the justificatory constituency
if they accept Gaus’s views on rational justification, this means Gaus’s approach (and by
implication the convergence model) does not apply to the world that we live in, nor does
it apply to a well-ordered liberal society” (272–3).
37
Dreben (2003) is a clear example of someone emphasizing both the ideal-theory
nature of the project (e.g., 323, 344) and the thought that important conclusions about
the real world follow (e.g., 317, 328). He doesn’t do anything to reconcile these arguably
inconsistent claims.
I think that Quong’s recent (2011) emphasis on the internal nature of the political
liberal project is in line with the Rawlsian ideal-theory way of excluding the unreason-
able (in conversation, Quong has confirmed this), and at times (e.g., 143) he even talks
explicitly in terms of ideal theory. Now, I think that Quong’s insistence on the internal
project cannot withstand criticism in general, but he develops this line in great detail,
and offers explicit justification for doing it. I remain unconvinced, and I hope to say
why in future work. Here, I must settle for noting that as a response to the problem of
excluding the unreasonable, Quong seems to fail for reasons similar to the ones in the
text here. See especially 290, where he seems to be guilty of precisely the fallacious
move from ideal to non-ideal theory highlighted in the text—having justified some-
thing under full compliance, proceeding to assume that it’s been justified to the com-
plying under partial compliance.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
38
See, for instance, Farrelly (2007), Stemplowska (2008), and Simmons (2010).
Unlike Simmons, I am not invested in any interpretive claim about how best to under-
stand these terms in Rawls. And notice also that the point I am about to make in the text
stands even if everything Simmons says (including about the priority of ideal theory, in
the sense he finds in Rawls) is true.
39
Quong seems to be officially okay with this result, but—as can be seen from the
references in footnote 37 above—doesn’t seem consistent in this regard.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
40
Again, see Dworkin (1973). For a recent attempt to respond to it—though in the
context of his specific, somewhat idiosyncratic, normative hypothetical consent theory—
see Estlund’s (2008) discussion of normative consent.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
41
“There may be no middle way between actual (including implied) agreement and
rational justification” (Raz 1990, 46).
42
This often-used kind of example comes, for instance, from Waldron (1987, 132).
43
See, for instance, Gaus (2011, Chapter 24).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
perhaps she’s not well informed; perhaps, while satisfying the minimal rea-
sonableness criteria, still she hasn’t been taking others’ point of view into
account sufficiently impartially. Because this is so, the public-reason theorist
reassures himself, there is really no legitimacy problem after all. All is well.
How convincing is this line of thought, as applied to real-life political
disagreements? Remember, we’re not talking about the highly anxious or
drunk patient, in the emergency room, in an especially vulnerable time in
his life. We’re talking about serious people taking part in serious discussions
over long periods of time, in the public sphere or in the privacy of smaller
social interactions, and anyway very deeply committed to their respectable
views. And let’s set aside for now other related worries, such as how civil it
would be to say to the one rejecting the principle that he would have
accepted it had he spent a respectable amount of good reasoning on it.44 The
question we’re interested in is whether such a response is in line with the
underlying public-reason motivations. If I am the person rejecting the sug-
gested taxation scheme, I am in the Rousseauean predicament—supposedly
free, but about to be coerced to take part in this policy which I reject. Is this
tension in any way reconciled seeing that I would have accepted the princi-
ple in some hypothetical conditions that are sufficiently far so that I am not
moved here and now, not even having thought about things quite seriously,
having debated them with my friends, and so on?
I am inclined, of course, to say that this is no response at all. The tension
between authority and liberty is fully present in these cases, even given this
kind of hypothetical justification-to. Now, let me emphasize again that I am
not claiming that hypothetical consent and justification-to never matter.
What I claim is that the thought that they matter here, where they are sup-
posed to matter most, in the context of political disagreement, resists belief.
Also, note that all of this is supposed to be from the point of view of a
public-reason theorist, one who thinks that justification-to is extremely
important. My point is not that it’s extremely important to justify the taxa-
tion scheme to all those rejecting it. My point is just that if you think that
justification-to is a necessary condition for legitimacy, because this is the
only way to reconcile authority with liberty and equality, or to respect as
equal and free your fellow citizens, then you should not be impressed with
the response at the end of the previous paragraph. That response amounts
to no progress in respecting your fellow citizens as free and equal. It should
not impress you more, for instance, than the response “I understand that
you are not convinced, but I’ve already stated my reasons for why it’s true
44
For this point, put in terms of adding insult to injury, and directed at Gaus, see my
(2013, section 6). Hampton (1993, 309) seems to anticipate something like this line. And
Wolterstorff (2012, 74) makes precisely this point as well.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
that this taxation scheme is justified.”45 And this response is, of course, not
one that a public-reason theorist can settle for without abandoning his com-
mitment to public reason.
Furthermore, the public-reason theorist who wants to use such idealiza-
tion owes us also the details—what the idealization does and what it does
not rule out. It’s not going to be easy to fill in the details in an extensionally
adequate way. And, what’s more, extensional adequacy is not enough. It’s
also necessary to offer a rationale for the specific way of idealizing, or the
specific hypothetical conditions deemed relevant, a rationale that’s consist-
ent with the underlying public-reason motivations. I don’t see how this can
possibly be done.
Going hypothetical, I conclude, is no more promising a strategy for the
public-reason theorist than excluding the unreasonable. Any such attempt
will be objectionably ad hoc. The public-reason theorist should either settle
for actual justification-to as a necessary condition for legitimacy (which in
all likelihood entails anarchism, even when restricted to just the reasonable),
or else take back his commitment to public reason.
45
A similar point comes up in the context of granting conscientious exemption from
military service. Statman (2009) argues that in the Israeli context pacifist objectors are
wrong about what follows from their own pacifist commitments, so that refusing to
exempt them is no attack on their conscience. I (2009b) respond by noting the point in
the text: The tension between freedom of conscience and the demands of the state is not
in any way relieved by the fact (if it is a fact) that the objectors are wrong about what
follows from their own deepest commitments.
46
This confusion pervades, I believe, Rawls’s and Rawlsian texts. For one clear exam-
ple, see Political Liberalism (1993, 129). Usually, though, the confusion is less explicit than
that. Still, attributing this confusion to Rawlsians is needed in order to make sense of
claims that equality is in any way relevant here.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
reason for action—that is, the feature of the circumstances that I take to
be normatively relevant—is not that I believe Catholic doctrine. Rather,
my reason is the content of that belief, namely Catholic doctrine itself.
Similarly, if I, a comprehensive, “metaphysical” liberal impose directives that
are based on the value of autonomy on people who do not value autonomy,
my reason for action is not that I value autonomy. Had this been the reason,
then given that they do not value autonomy, privileging my valuings over
theirs would have violated some intuitive principle of equality. But my reason
for action is different, and is not about me at all. Rather, my reason for action
is that autonomy makes people’s lives go better, or some such. There is nothing
indexical about this reason for action, and so no violation of equality at all.
The point is tricky, and so merits more elaboration. It’s tricky partly
because it’s hard to distinguish, from the first-person perspective, between
that-p and that-I-believe-that-p. But with the help of some counterfactuals,
this can be done. We’re assuming, then, that in the actual world, you think
that autonomy is of value. Let’s imagine a possible world in which you are
mistaken about autonomy, thinking that it’s not of value, even though it is.
Do you still want—in the actual world—to impose autonomy-based direc-
tives on people in that hypothetical world? A positive answer seems obvious.
But in that world, you don’t believe in the value of autonomy. Still, in that
world autonomy is of value. So what we can learn from the fact that when
you think (in the actual world) about that possible world you still want to
impose autonomy-based directives is that your reason for imposing them is
that autonomy is of value (which is true there too), not that you believe that it
is of value (because on that world, you don’t). Compare this to cases of, say,
taste. Suppose that in the actual world you prefer vanilla ice cream to choc-
olate. Now imagine a possible world in which you prefer chocolate. And
suppose that you—in the actual world—have to order ice cream for your
counterpart in that possible world. What flavor do you order? Clearly, choc-
olate. What this shows is that in the actual world too, when you order
vanilla ice cream, your reason for action is indexical, it is about you and your
preferences, it’s simply that you prefer vanilla. This is why it makes no sense
to order vanilla ice cream for your counterpart in a world in which he no
longer prefers vanilla. But when you act on a principle or something like the
value of autonomy, this simply isn’t so. Your reason for action in such cases
is that-p, not that-you-believe-that-p. It’s just not about you at all. So, there
is no sense in which you’re giving extra weight to your beliefs over others’—
you’re giving no weight to your beliefs here. You’re just giving weight to the
value of autonomy, or the truth of Catholicism, or some such.
This point is in no way special to the political realm. It applies across a
wide range of cases, in practical as well as theoretical reasoning, regarding
justification and explanation alike. Suppose you believe that what explains
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
the tide is the gravitational force of the moon. What, do you believe, does
the explanatory (and causal) work here—that the moon has this-and-that
gravitational force, or that you believe that it does? Clearly, the former. You
don’t believe that your beliefs causally regulate the tides, nor do you think
they explain them. What has relevant explanatory force is the content of
your beliefs, not the fact that you have them. We can run the counterfactual
test again—surely, when you think (in the actual world) about a possible
world in which you’re wrong in your beliefs about the explanation of the
tides, you don’t (in the actual world) think that the explanation of the tides
in that possible world is any different than the one in the actual world.
What does the explanatory work is entirely present there—it is the moon’s
gravitational force, not your beliefs about it (which are absent there). The
explanation, then, is not about you at all.
Similarly, in epistemology the question sometimes arises whether—in a
case of disagreement with someone you take to be your epistemic peer on
the relevant question—you should give extra weight to your own evalua-
tion of the relevant evidence. Here too, though, it’s clear that in typical
cases, the thing to which you accord epistemic weight is not that you eval-
uate the evidence a certain way, but rather the evidence and its epistemic
value. You are not a part of the picture of what does the epistemic work. It’s
just not about you.47
And the same point applies to our reasons for action even when they are
not themselves normative. If I believe that cutting the blue wire will defuse
the bomb, you believe that cutting the red one will, and I proceed to cut the
blue one, there is no sense in which I am giving extra weight to myself or my
beliefs. My reason for action is not that I believe that cutting the blue wire will
defuse the bomb. This is not the feature of the case to which I accord norma-
tive significance (as can be seen, for instance, by running the counterfactual
test again). Rather, my reason for cutting the blue wire is that doing so will
defuse the bomb. And this reason is not indexical at all. Once again, then, it’s
not about me.48
47
This is one of my main points against the Equal Weight View in “Not Just a
Truthometer” (2011b). For similar points in a wider epistemic context, see Schroeder
(2008). Of course, not many things are uncontroversial in the peer disagreement litera-
ture. But the point in the text, it seems to me, should be. (Those I criticize in “Not Just a
Truthometer” do not argue against this point; rather, they seem to either assume its denial,
or not to see its full significance.)
48
You may be worried about mistakes: what if in fact autonomy is not of value? Surely,
in that case my reason for action can’t be that autonomy is of value (for it isn’t), and the only
remaining candidate seems to be my belief in the value of autonomy. But this is not so. A
fuller discussion of these issues—like the one I offer in (2011a, 221–3; see also the references
there)—will distinguish between different senses of “reason for action,” only some of which
are factive. In the terms I use there, the agent’s reason—the feature of the situation the agent
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
Now, you may think that still there is a sense in which I am not out of the
picture when it’s the content of my beliefs that plays the relevant role in my
reasons for belief or action. And I think that this is right, but we have to be
careful about the precise nature of this role.49 The point is, again, that the
fact that it is my view is not a part of what is, on my view, normatively rel-
evant (unlike, for instance, in the mere preference for vanilla case). It is thus
not a part of my reason for action. It may be a necessary background condi-
tion for my reason being a reason for me to act in the relevant way. But this
is very different:50 something about me may be a part of the story of why I
offer the value of autonomy as a reason for (political) action. But it’s not a
part of the reason itself. Again, the situation in other cases is exactly alike:
perhaps the fact that I believe that the moon’s gravitational force explains
the tides is a part of the background conditions needed for my offering that
explanation of the tides. But this belief of mine is not a part of the explana-
tion itself. The explanation is not about me.
Still, you may want to insist, even if nothing about me is a part of my
reason for action, still I am involved in this other, background-ish kind
of way. Furthermore, even if I do not, by acting politically on the value of
autonomy, give extra weight to my belief about the value of autonomy
(because that belief is no part of my reason for action), still I act in a way that
will, as I know, bring about a situation in which my beliefs are more in line
with the state’s actions than the beliefs of those rejecting the value of auton-
omy. In this sense too, then, it may still be argued that acting on the value of
autonomy, in the face of (reasonable?) disagreement about it, violates equal-
ity. But this would be a mistake. It’s true that when I act, I act on my own
beliefs. There is no way around it: anything I do is something that I do. Even
if I decide to defer to another, it’s me doing the deferring, and indeed the
deciding to defer. Even if I become a card-carrying member of the public-
reason club, the refusal to act politically on my comprehensive doctrine will
be very much my refusal. My beliefs and actions are and always will be my
beliefs and actions.51 Thus, if by acting on the value of autonomy I violate
equality (because it’s my own beliefs I act on), then any action whatsoever
violates equality. This is not the kind of equality worth caring about.
finds normatively relevant—need not be factive. Even in the case in which autonomy is not
of value, then, my reason for action may be the value of autonomy (just like even if the
moon is not after all responsible for the tides, still what in my view does the explanatory
work is the moon’s gravitational force, not my belief about it).
49
Raz (1998, 27) puts this point correctly but somewhat cryptically thus: “We rely on
the answer to the question not because it is our view, but because it is, as we believe, true.”
50
This distinction is central to Mark Schroeder’s work. See, for instance, his (2007,
Chapter 2).
51
See Raz (1998, 27).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
What the discussion in this section shows is that the equality underlying
public-reason motivation is based on a simple confusion. The liberty motiva-
tion should be, to an extent, accommodated (a point I get to below). But the
equality motivation should just be rejected. The tension between authority
and equality (as understood by public-reason theorists) is a pseudo-problem.
52
See, for instance, Raz (1990, 10).
53
See Hampton (1989). Rawls’s (1989) is motivated in large part by the need to
respond to this accusation.
54
Rawls (1993, 10).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
55
There may be more questions still. One (i) is the first order question, on its own.
Another question (ii) is the one in the text—about how to deal with disagreement.
Possibly there’s a third one—revisiting the first order question, now with the answer
to the second one in mind; roughly this is the question (iii) whether to intervene in
Syria given the disagreement, or taking into account also the fact that people differ
about (i). One doesn’t have to be a public-reason theorist to acknowledge a difference
between (i) and (iii). Disagreement may be one normatively relevant fact here, among
others. (And once we see this, we may also ask many more questions: (iv) how to
respond to (iii) given that there’s disagreement about it as well; (v) how to answer (i)
given disagreement about (i) and (iii); (vi) how to answer (v) given disagreement
about it; . . .).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
56
Not just this quote comes from Raz (1998, 47), but also the thought expressed here
more widely: it echoes Raz’s insistence that treating people with respect requires engaging
them, not, for instance, propagating false beliefs (1998, 43). I think that Raz overstates
his case—respecting people amounts to different things in different contexts, and per-
haps sometimes is best manifested by a refusal to engage them. Not, though, in the
context in the text, certainly not in the patronizing public-reason kind of way.
57
For a close point see Raz (1998, 35, footnote 13).
58
Nagel (1987, 158) does notice this, as when he says: “Impartiality among persons is
one thing, but impartiality among conceptions of the good is quite another.” But at the
end of the day he thinks that this gap can somehow be bridged. In another (metaethical)
context, I discuss in detail the distinction between conflicts that are based on mere pref-
erences and those that are based on moral disagreement. See Chapter 2 of my Taking
Morality Seriously (2011). There is some similarity between the point in the text and an
oft-made criticism of political liberalism, according to which it prevents (for instance) the
religious from relying on their deeply held beliefs in the public domain. But the point in
the text is different—it’s not about what considerations political actors are allowed to cite
in the public domain, exactly, but about how the public-reason theorist thinks of his own
role vis-à-vis such debates.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
Nor, it seems to me, is it what it has been like in the field’s long history (most of
60
61
I don’t know of anyone pursuing this line, but there are some related hints in the
literature. The one most clearly flirting with this line is Brower (1994, 22). And there
seems to be something in this direction also in Hampton (1989, 801), Nagel (2003, 77),
Estlund (2008, 163), and Kelly Sorensen’s (2013) “aspirational political liberalism.”
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
62
Similarly for the self-defeat worry (again, see Wall (2002), and my (2013)). And I
am pretty sure that this is true also of the epistemological problems facing public-reason
accounts, but I have to fill in all the details there.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
Bibliography
Brower, Bruce (1994). “The Limits of Public Reason,” The Journal of Philosophy 91,
5–26.
Burton, Dreben (2003). “On Rawls and Political Liberalism,” in Freeman, 316–46.
Cohen, Joshua (2009). “Truth and Public Reason,” Philosophy and Public Affairs 37,
2–42.
Cuneo, Terence (2013). Book Review of Paul Weithman’s Why Political Liberalism?,
Faith and Philosophy 30, 357–61.
Dworkin, Ronald (1973). “Justice and Rights,” in his Taking Rights Seriously
(Cambridge: Harvard University Press), Ch. 6.
Enoch, David (2005). “Why Idealize?,” Ethics 115(4), 759–87.
Enoch, David (2007). “Taking Disagreement Seriously: Some Critical Comments
on Jeremy Waldron’s Law and Disagreement,” The Israel Law Review 39, 22–35.
Enoch, David (2009a). “On Estlund’s Democratic Authority,” Iyyun 58, 35–48.
Enoch, David (2009b). “More on the Conscience-based Exemption from Military
Service: A Reply to Dani Statman,” 31 Iyunei Mishpat (Tel Aviv University Law
Review), 709–40 (in Hebrew).
Enoch, David (2011a). Taking Morality Seriously (Oxford: Oxford University Press).
Enoch, David (2011b). “Not Just a Truthometer: Taking Oneself Seriously (But Not
Too Seriously) in Cases of Peer Disagreement,” Mind 119, 953–97.
Enoch, David (2013). “The Disorder of Public Reason: A Critical Study of Gerlad
Gaus’s The Order of Public Reason,” Ethics 124, 1–36.
Enoch, David (2014) “Authority and Reason-giving,” Philosophy and Phenomeno
logical Research 89, 296–332.
Enoch, David (Manuscript). “Idealizing Still Not Off the Hook,” available at
<http://law.huji.ac.il/upload/ReplyToSobel.doc>.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
PA RT I I I
RIGHTS AND DUTIES
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
6
Territorial Rights
Justificatory Strategies
A. John Simmons
States are defined in international law (for instance, in the 1933 Montevideo
Convention) as entities with permanent populations and fixed territories
under government control. I will focus here on just part of that definition: the
idea of states as territorial entities (indeed, as necessarily so, from this legal
viewpoint). Most of us do think of states in terms of their territories, first
learning about our own and other states by locating their colored territorial
shapes on maps. Exactly where the territorial lines defining these shapes are
drawn is obviously a matter of considerable importance to states, and the
history of the conflicts in which modern states have engaged has regularly
involved attempts, often successful, to redraw those lines by force.
Each modern state identifies itself with and claims a set of exclusive rights
over a particular territory.1 It is commonly assumed, I think, that these
claimed territorial rights are not merely legal rights; they are, at least in
decent or legitimate states, also moral rights over territories (or morally jus-
tified legal rights). It is on the possible moral justifications of states’ territo-
rial rights that I concentrate here. These rights include at least the right to
exclusive legal jurisdiction over the territory (that is, the right to coercively
regulate the conduct of all within the territory by means of enforcing all
legal rules and directives of the state); the right to full control over the land
and resources within (or constitutive of ) the territory that are not privately
1
I intend by this to deny neither the possibility of federal governance within states
nor the possibility of voluntary renunciation, transfer, or collective exercise of rights by
states. The rights discussed below should be thought of as those to which autonomous
nation-states take themselves to be entitled, prior to or independent of any voluntary
reduction or sharing of those rights within or between states.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
owned; the right to tax and regulate privately owned land and resources
within the territory; the right to control or prohibit movement (of persons
and materials) across territorial boundaries; the right to determine the
standing of those within the territory (by, e.g., establishing rules governing
residency, diplomatic status, or citizenship); and the right to prohibit indi-
vidual or group territorial secession or alienation of territory to nonmember
persons or groups.2
Some of these claimed rights, notice, are primarily jurisdictional in
nature—that is, they are claims primarily to regulate and control a particular
territory, rather than claims to exclude persons from entering or using it.
Others of the territorial claims states make are more property-like, exclusion-
ary claims over a region, claims to choose who or what may pass over its
boundaries and who may use (and how they may use) the resources located
in it. I will not here discuss these more property-like claims, concentrating
instead on states’ jurisdictional claims over the territories with which they
identify themselves.3 The jurisdictional claims that states make are not, of
course, merely claims they make with respect to the geographical space
itself; they are primarily claims over the human beings located in that space.
States claim authority over those within their territories, the rights to make
and enforce laws and directives for them and the right to peaceful compliance
from them. States claim the right to “speak the law” to a set of persons—as
the literal meaning of “jurisdiction” suggests—and the relevant set of persons
is in the modern world primarily located in territorial terms. As a result, the
word “jurisdiction” now refers as well to the geographical area over which
legal authority is thought to extend.
I should emphasize that I am concerned here with the justification of
states’ claims over their particular territories, rather than with other kinds
of possible moral claims concerning group or state control over land and
resources (or about the modern state’s territorial form). For instance, it has
been regularly argued both that certain kinds of groups of persons have by
their natures a right to a territory on which to govern themselves—even
when there is no particular piece of land to which they obviously have a
valid claim—and that it is best in various ways for all or most of the land in
the world to be controlled by states or societies, rather than to be simply
open for all to use—even when there is no special reason why one state
rather than another should control any particular portions of the earth.
2
This is intended only as a list of the principal rights actually claimed by modern
states, not (as Cara Nine appears to suppose) a list of the necessary elements “of the con-
cept of territorial rights” (Nine [2012], 11).
3
I believe that the property-like rights modern states claim are in fact much harder to
justify than are their jurisdictional claims.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
2. STRATEGIES
4
E.g., “If we consider the range of functions that modern states perform, it quickly
becomes obvious that these functions cannot be carried out effectively unless the state has
authority over a determinate territory” (Miller [2007], 214).
5
In Chaim Gans’ nationalist theory of territorial rights, the distinction I make here is
identified as the distinction between “the right to territorial sovereignty” and “the loca-
tion of territorial sovereignty” (Gans [2003], 103).
6
Most of these theories are quite recent. When I first explored the topic of territorial
rights in print (in 2001), almost nothing had been written on the subject by contempo-
rary thinkers. The great historical political and legal philosophers also had very little to
say about such rights.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
7
E.g., Altman and Wellman (2009).
8
Simmons (2001).
9
This is Anna Stilz’ name for them in Stilz (2011), 576. See, e.g., Buchanan (2004),
370–1; Christiano (2006).
10
Stilz (2011), 581–2. Stilz’s basic position is Kantian; but her final account of states’
territorial rights, as we will see, is actually more of a hybrid.
11
The first apparent defense of a utilitarian theory of territorial rights is in Sidgwick
(1908), 252.
12
E.g., Gans (2003); Meisels (2009); Miller (2007).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
13
See, e.g., Pogge (1992), (1994). Left-libertarians are also generally skeptical about
the territorial claims made by most actual states; but I will not here discuss their position.
For a very brief summary of a left-libertarian view of territorial rights, see Steiner (2005),
34–6.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
prove useful to identify the virtues and limits of the pure forms. For in
doing so we will be locating the argumentative “cores” of those attempts;
and we will then be better able to see the directions in which the failures of
the pure forms have pushed those attempts—and the directions in which
those attempts may need to be further pushed in order to successfully justify
strong territorial rights for actual states.
3. NATIONALISM
14
Miller (2007), 217–18.
15
Gans (2003), 116.
16
Miller (2007), 219.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
territorial rights simply may not apply to many actual decent states, threatening
to yield only skeptical conclusions.
Especially troubling are cases in which the state currently occupying (and
claiming) the territory is not the one—or the only one—with the appro-
priate sorts of historical and cultural connections to the land to trigger
nationalist-style reasoning about territorial rights. And, of course, these
cases are most troubling when the current occupant took possession of the
territory in question by plainly unjust or illegitimate means. Such cases are
unhappily commonplace, so any nationalist account that hopes to apply its
arguments (non-skeptically) to the real world is obliged to address them.
The standard argumentative move is to simply claim that the rights of inno-
cent peoples (and persons) who are illegitimately annexed, conquered, or
expelled “fade out” with the passage of time, while new rights for those who
have wrongly seized their territories (or for their descendants) “fade in.”
While it is, of course, hard to be very precise about this process, it is a pro-
cess that is assumed by many (including non-nationalist) writers on the
subject to reflect the moral facts. Miller, for instance, denies that his posi-
tion amounts to “a charter for thieves,” but allows that the question of who
eventually “has the better title will be a matter of judgement.”17 So while the
wrongs states and groups do in achieving their territorial goals may initially
call for restitution, for a restoration of the status quo ante, once those wrongs
become older, the wrongs are “superseded”18 and there springs into being a
new set of moral rights for a new set of people to control the unjustly
acquired territory.
Because virtually no modern states can make territorial claims that are not
historically stained by such injustices, it may seem that a view like Miller’s
is a necessary feature of any adequate theory of territorial rights. We should
note two points, however. First, the devil here is surely in the details. Exactly
when and why rights go away and appear, how soon and for what reason
victims lose their claims to restitution or reparation, and wrongdoers (or
their heirs) acquire claims to ill-gotten gains, is a matter of significant theo-
retical and practical importance and great theoretical difficulty. Second, we
will only be required to accept a very dramatic and suspiciously self-serving
account of the supersession of our states’ territorial wrongs if we think a
standard of adequacy for theories of territorial rights is that they not be
interestingly revisionist in their implications—that is, that they be able to
explain why stable, reasonably just states in the world actually have legiti-
mate territorial jurisdiction over all of the territory that they claim as theirs.
If we are prepared to accept instead the possibility that even reasonably just
17
Miller (2007), 220, 219.
18
Waldron (1992).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
states may not be morally entitled to all they claim, we can defend a more
plausible account of the moral significance of the many relatively recent
wrongs done by states in their quests for territory. For now, however, I want
simply to flag this problem confronting nationalist accounts—call it “the
problem of rights supersession”—since it is, I think, both a serious problem
and a problem that (as we will see) also confronts many non-nationalist
accounts of states’ territorial rights.
Surely, though, the most severe hurdle faced by nationalist accounts of
territorial rights is simply the absence of cultural/national uniformity within
the marked boundaries of the political world. Nationalists tend to locate the
relevant territorial rights in the cultural or national majorities within plural-
istic political units, leaving the preferences, interests, and goals of those not
included in the majority national group disturbingly to one side. Miller’s
response to the problem is to claim that while all residents of a national
territory may not share all of the same goals, they all do have “a common
interest in being able to set those goals through democratic debate” and in
belonging “to a community with a shared sense of national identity.”19 But
this response seems only to push the relevant difficulty one step back.
Minority groups and individuals may well have a version of the interests
described by Miller. But they might well have no interest at all in a shared
setting of goals, engaging in democratic debate, and achieving a shared
national identity in the particular political setting in which they find them-
selves—a setting in which they will be systematically outvoted on matters of
central concern to them by the majority national group. Call this difficulty
“the problem of trapped minorities.” It is again, as we will see, a difficulty
shared by a variety of non-nationalist approaches.
4. FUNCTIONALISM
happen to arise in one place or another need not reflect any special relation-
ship between those states and the land they occupy. If those states could
function effectively elsewhere or function effectively with altered bounda-
ries, simple considerations of their morally mandated functions present
no principled bar to such changes. Without additional, non-functionalist
principles in play, it is hard to see why our relocating, adding to, of sub-
tracting from a legitimate state’s territory would constitute a wrong. Further,
of course, currently functional states can plainly rest on a sordid history of
wrongdoing. What matters for the functionalist is that the state in ques-
tion here and now successfully administers justice or successfully promotes
social happiness.
This means, of course, that functionalist theories will not only have prob-
lems with the particularity of territorial claims, they will also face the same
problems of trapped minorities and rights supersession that face nationalist
theories. States can perform their morally required functions even with
unwilling parties and groups subjected to political authority within their
borders; and the functionalist, like the nationalist, owes us a convincing
account of when and why the rights of wronged parties just fade away as
they are opposed by the interests of established, adequately functioning
states. These difficulties for functionalism add up to what I have elsewhere
called the “the boundary problem.”20 Pure functionalism simply cannot
guarantee that the boundaries of the territories counted by the theory as
rightfully subject to state authority do not enclose people who have plainly
been subjected illegitimately to states’ coercive power. Because functional-
ism grounds states’ territorial rights in their current provision to areas of the
services that states are morally required to provide, states can acquire justi-
fied territorial rights, according to the logic of the functionalist approach,
simply by making themselves the provider to a territory of those services.
Suppose the United States one night secretly moved its southern border
barriers a few miles further south into Mexico, claiming the newly enclosed
Mexican territory and the surprised Mexican subjects living on it as its
own.21 There would appear to be nothing wrong with this according to
functionalist reasoning—provided only that the United States extended its
effective administration of justice to this new territory as well. It is a state’s
effective administration of justice over a territory and people, not the history
of the state’s acquisition of territories and subjects, that functionalism iden-
tifies as the source of its legitimate territorial rights.
Simmons (2013).
20
I am supposing further, for purposes of this example, that both states are sufficiently
21
just (or possess whatever other legitimating properties are necessary) for the relevant
functionalist view to count both as having started with legitimate territorial rights over
their claimed territories.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
22
There is, however, no very obvious reason why nations could not form bizarre
attachments to (or otherwise build their identities around) distant territories or land-
marks with which they had little or no actual history of physical interaction.
23
Altman and Wellman (2009), 4–5.
24
Altman and Wellman (2009), 46–7.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
25
All are free to exit, of course, but only by abandoning their lands and subjecting
themselves to the dominion of whatever other society (if any) is willing to take them in.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
26
See Nine (2012), 58.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
27
Altman and Wellman (2009), 49.
28
Altman and Wellman (2009).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
This seems to me, however, a far deeper and more serious problem than
their response acknowledges. Both the boundaries of the original states and
the boundaries picked out by those forming new states may be morally
arbitrary (or deeply illegitimate) in very important ways. For instance, “the
separatists” discussed by Altman and Wellman might easily include in their
“specified” territory not just the land occupied by their supporters, but also
land occupied by others (who will be easily outvoted by those supporters),
possibly because that land contains valuable resources or other desirable
geographical features—just as the territory from which they desire to sepa-
rate may itself have been formed by “trapping” unwilling minorities in var-
ious ways. Provided only that this “specification” does not incapacitate their
original state (or take an unjust share of the state’s resources, say), there
appears to be no bar in the theory to such majority choices by secessionists.
While Altman and Wellman do condemn the annexation of less popu-
lous by more populous states,29 they do so only where the less populous
groups are already organized as states. Those people and groups who are
“unable” to function as states (for whatever reason?) are simply left as grist
for the statist mill.30 Altman and Wellman seem primarily concerned to
affirm that unwilling individuals, sprinkled here and there throughout an
otherwise willing and territorially contiguous group, may be legitimately
subjected to political authority without their consents. But the wrongs their
position permits—in both terms of trapped minorities of significant sizes
and rights superseded in an instant in the interest of the present possessor of
territory—seem vastly more significant than those they seem principally
concerned to deny. And these difficulties push the theory in a more fully
voluntarist direction, toward a view capable of greater sensitivity to histori-
cal wrongs against peoples and persons.
29
Altman and Wellman (2009), 52.
30
Wellman does elsewhere insist that “willing and able” political groups “who occupy
a territory enjoy jurisdictional rights over this land” only “other things being equal.” But
the only example he gives of a case where other things are not equal is again that of the
forcible annexation of one state by another (Wellman and Cole [2011], 102–3).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
those around them, those whose arrangements they are may not interfere
with the rightful activities of the unwilling.
Similarly, Lockean voluntarism rejects the idea of simple supersession of
rights by seizure or mere passage of time (moral rights being, on this view,
“imprescriptible”).31 It cannot be embarrassed by theoretical insensitivity
to the plights of the expelled, the annexed, and the wrongly subjected—
even when those unfortunates are mere individuals or when, as groups, they
have never been able or been permitted to organize politically. The Lockean
view is thus not vulnerable to charges of over-eagerness to simply affirm
the legitimacy of the territorial status quo. The rights of those maimed in
the machinery of politics do not simply fade away for the convenience
of the powerful or the numerous. Those whose rights have been violated in
creating or reshaping states retain the right to rectification of those wrongs.
Rights supersession is, according to the Lockean voluntarist position, nor-
mally just wishful thinking, typically done by those who most stand to
benefit from it.32
With all of these noteworthy virtues, why have political philosophers and
political theorists not jumped at the opportunity to embrace the Lockean vol-
untarist theory? The answer is not, I think, that proponents of non-Lockean
views simply reject the powerful intuitions on which the Lockean theory is
built. For it is at precisely the point where concerns about historical wrongs are
brought clearly into focus that most rival theories are motivated to take hybrid
form, introducing historical principles in order to avoid the strongly counter-
intuitive implications of their core theories. I take, instead, the most persistent
and fundamental objections to Lockean voluntarism to be three.33 The least
31
Lockean voluntarism must still develop a theory of “moral residue” for cases where
rightholders and wrongdoers have died with neither rectification of the wrongs nor
inherited rights and duties of rectification. I know of no satisfactory theory of that sort.
But all theories of territorial rights are equally in need of one.
32
Most Lockeans (including myself ) hold, as well, that rights over land are heritable
and do not simply disappear as the specific rightholders and wrongdoers die off. And
rights over land held by freely incorporated groups continue to be held by those groups,
even as their memberships change.
33
I leave to one side an equally common basis for objection—namely, the defects in
the arguments advanced by Locke himself. Stilz, for instance, appears to (mistakenly)
reject the Lockean approach to territorial rights almost entirely because she (correctly)
takes Locke’s own arguments to be defective. Locke’s conservative ambitions—to justify
continuous, enduring rights over their full claimed territories for non-tyrannical states—led
him to compromise his own foundational principles and to make some of the errors Stilz
identifies (Stilz [2009], 192–4). But a more genuinely voluntaristic (and less conservative,
more revisionist) Lockean position than Locke’s own, one that takes seriously each per-
son’s equal right to a fair share of the earth and its resources, is surely a better test of the
Lockean approach to these issues—just as Kantians (like Stilz) often find their own posi-
tions more defensible when they depart from the letter of Kant’s texts (Stilz [2009], 198
[note 8], 203).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
troubling of these, despite its surprising resilience, is the claim that the Lockean
account confuses the idea of property with the quite different idea of juris-
diction or territory, in consequence of which the Lockean account of states’
territorial rights must be equally confused.34 To identify the two would, of
course, be confused. But property ownership clearly has a jurisdictional
aspect, just as territorial rights have a property-like aspect (making it not at
all confused to suppose that the latter might derive from the former).
Landowners have (“jurisdictional”) rights to make certain kinds of rules
to govern their lands, thus unilaterally restricting the liberty of those who
choose to enter on their land. Landowners who choose to submit their land
to a state’s authority give the state’s rules priority on their lands, thus
accounting for the state’s consequent jurisdictional rights over those lands.
They also agree to allow the state to regulate their land in other ways (which
they were originally entitled to do themselves), including controlling those
boundaries of it that will constitute parts of the state’s territorial boundaries.
The result is a kind of sharing between state and subject of the various inci-
dents that comprise full rights in land, and a concentration of some of those
incidents (received from all subjects) in the hands of the state. This latter
concentration, I think, is an accurate representation of our normal concep-
tion of states’ rights over their territories, not any kind of confusion of ter-
ritory with property.
Often this first objection is combined with a second, in whose light the
first becomes more intelligible. If rights in land are necessarily created
through legal or institutional rule, then it plainly makes no sense to claim
that individual rights in land serve as the justifying foundation for these
institutions (i.e., the institutions that are responsible for creating rights in
land in the first place). The first objection thus really appears often to rest
on the second: that it is not possible to make sense of rights in land outside
of a state whose laws establish them; so we naturally can make no sense
either of grounding a state’s territorial rights in the subjection (by willing
members) of private rights in land to that state’s jurisdiction. Notice, how-
ever, that the required skepticism about pre-legal rights in land must extend
further than mere skepticism about Lockean “natural” property rights. It
must extend as well to all conventionalist theories of rights in land (such as
Hume’s), according to which extra-legal moral rights in land and chattels are
possible given the establishment of appropriate interpersonal conventions
to define and regulate them. If extra-legal property in land is possible—be
it naturally or conventionally grounded—then such land can be subjected
to a state’s jurisdiction, as the Lockean model requires for justified territorial
rights. So this objection must claim quite a lot about the impossibility of
private rights in land outside the state, requiring somewhat more philo-
sophical nerve than at first might seem to be the case.
The third and undoubtedly most fundamental obstacle to the acceptance
of Lockean voluntarism, however, is that the theory offers an account of
states legitimate territorial rights that does not match up very well with the
real-world claims made by actual modern states.
States are not voluntary associations, nor have the territories states’ claim
been constructed from the submission of land by willing subjects to state
jurisdiction. Lockean theory thus seems to describe an extremely demand-
ing ideal that is simply too distant from the hard reality of the world to
permit its application to real-world territorial claims and disputes.35 The
plebiscitary voluntarist, functionalist, and nationalist accounts may appear
to do better on this score, since they are focused more on states’ present
characters and capacities and less on the historical processes that produced
the present distribution of states’ claimed territories. These views thus seem
more immediately applicable to real-world territorial claims and disputes,
despite their shared theoretical defects (described above), and less likely to
require dramatic revisions in our views about the actual territorial rights of
real-world states.
In part, this complaint about the Lockean theory is simply correct.
Because it takes very seriously historical wrongs and the rights of the unwill-
ingly subjected, the theory’s practical implications will inevitably be more
revisionist than will be those of the alternative theories I’ve criticized here.
But all of these theories are ideal theories; all describe ideals to which real
states’ territorial claims will conform only very imperfectly. Many real states’
territorial claims will not qualify as legitimate on the terms of plebiscitary
voluntarism, functionalism, or nationalism, nor will these theories obvi-
ously yield clear solutions to the most pressing of the world’s actual territo-
rial disputes. These are, after all, philosophical theories.
The job of philosophical theories of this sort, in my view, is to identify
our moral target, to describe how states’ territorial claims could be fully
rightful and legitimate. There will then be separate and difficult questions
about how, in a distinctly non-ideal world, we can approach that moral
target in an efficient and fair fashion. That will involve, first, determining to
what extent territorial claims made by real-world states count as legitimate
according to the ideal theory; and, second, using the ideal to identify the
most serious wrongs done by states (and their subjects) in establishing their
acknowledged territorial boundaries. Only then can we meaningfully proceed
35
Miller, for instance, rejects such views as “impossibly demanding . . . , putting virtu-
ally all borders into question” (Miller [2007], 220).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
with the business of recommending practical policies that best correct these
wrongs, beginning with the most serious, and always focused as well on
practical and moral limitations on required reforms.
The true test for an ideal philosophical theory, in my view, is not how
closely and comfortably its prescriptions match the ways in which we actu-
ally live our political lives, but rather how plausibly it identifies the most
grievous kinds of wrongs that we do each other in the course of those lives.
And Lockean voluntarism, I submit, identifies the wrongs that need right-
ing in a clear and compelling way, putting us in position to attempt to
redress them and to gradually achieve a more rightful condition. For the
Lockean, our political ideal must be a world in which each person is, as fully
as possible, treated as a self-governing equal.
The alternative theories I’ve criticized are theoretically insensitive to too
many of the clear wrongs that states do in our names. When states insist
on exercising jurisdictional authority over land occupied by unwilling
subjects, or when they expel or destroy the innocent in their quests for
territorial control, they wrong persons in ways that require rectification,
either through the adjustment of territorial boundaries or through genu-
ine reparations of some other sort. Indeed, I think that modern states also
frequently do wrong simply in their efforts to control the sites of a wide
range of the world’s natural resources (through their more property-like
territorial claims).
In real-world disputes about territorial boundaries or about rights to nat-
ural resources, facts will inevitably be contested, claims will inevitably con-
flict, and disagreement will inevitably persist. A sound philosophical theory
of territorial rights can at best only identify the salient moral vectors in play
in such disputes and lay out the general guidelines for pursuing the best
resolutions of them.
7. ALTERNATIVES?
36
I defend this view of the relationship of non-ideal to ideal theory in Simmons
(2010a).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
8. PLURALISM
As I argued above, the standard reasons for hasty rejection of the Lockean
voluntarist position—on both territorial rights and political authority—are
less compelling than they might initially seem. Can nationalist, functional-
ist, or plebiscitary voluntarist theories effectively avoid these perceived
“costs” of embracing Lockean voluntarism by dealing in other ways with the
problems the Lockean view allows us to avoid, such as the problems of
trapped minorities, and of awkward commitments regarding the superses-
sion of rights? Perhaps those theories could simply be converted into more
pluralistic, hybrid accounts by adding a set of new principles designed to
correct their core theories’ inabilities to effectively address these problems. I
will here briefly examine just one recent effort to “hybridize” an alternative
theory of territorial rights: namely, Anna Stilz’s Kantian theory of territorial
rights. The problems with Stilz’s account are, I think, representative of those
facing hybrid accounts generally.
According to Stilz, a state enjoys “a prima facie claim to territorial juris-
diction” over a particular geographical territory if, first, “the state is in fact
legitimate” (i.e., it protects “freedom-as-independence to a sufficient degree”)
and, second, the state’s “legal system defines property rights over [that] par-
ticular area of land.”37 The “extent” of states’ jurisdictions and the particular
“locations” of their territorial rights thus at first appear to be determined
simply by the practical “reach” of legitimate states’ institutions for adminis-
tering justice. Legitimate states are morally entitled to govern whatever par-
ticular territories they actually effectively govern. So far, this is just the kind
of account that we see in other Kantian accounts of territorial rights. After
all, what appears to matter centrally in the Kantian project in political phi-
losophy is that freedom be secured by the subjection of all to effective jus-
tice-administering institutions, not that individuals be subject to this or
that particular political/legal administrative structure. So the problem of
identifying the particular territories over which states have de jure territorial
rights is allowed to be settled according to the location of the de facto terri-
torial control exercised by legitimate polities.38
Stilz, however, correctly recognizes that “if this were all that mattered . . . there
would be no objection to annexation,” since the annexing state might be
able to administer justice as well as (or better than) the existing domestic
view in contemporary political philosophy: the boundaries of existing states are “acciden-
tal” or “arbitrary,” but they are not so in any way that challenges their legitimacy. See, e.g.,
Rawls (1999), 39; Dworkin (2011), 381–2.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
39
Stilz (2011), 595.
40
See Simmons (2010b), (2013).
41
Stilz (2011), 590.
42
Stilz (2011), 583, 585. It thus remains unclear from Stilz’s discussion how, if at all,
we are to explain the (apparent) wrong involved in coercively relocating faultless people
whose goals, projects, and so on are not specially connected to the place in which they
legally reside. In her more recent work on occupancy rights, it appears that the only
wrong at issue in such cases may be the generic wrong of coercing people without justifi-
cation (Stilz [2013], 340–1). But if relocation does not involve violating an occupancy
right, justifying relocation may turn out to be quite easy. On the Lockean view, by con-
trast, persons have enduring rights over all of the land (within their fair share) of which
they make productive use.
43
Stilz (2011), 591, 595.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
44
Stilz (2011), 584.
45
Stilz (2011), 586.
46
Stilz (2011), 585. It is not immediately obvious why even the “fault” of the original
aggressive displacers should not also be capable of being “superseded” on an account like
Stilz’s. Her acceptance of the supersession of the victims’ rights seems to grow almost
entirely out of her concern that once the new occupants are “settled” in the territory, it
will be “impossible to move [them] without damage to nearly all [their] life plans” (584);
this is “what really counts for supersession” (583). But if wrongful aggressors manage to
refocus their life plans quickly enough, shouldn’t they also acquire rights of occupancy
(that supersede those of displaced parties) in this way? Stilz has more recently added the
further requirement that one can acquire an occupancy right only if “his connection to
the territory was established without any wrongdoing on his part” (Stilz [2013], 354). But
if historical wrongdoing does, indeed, in this way limit the acquisition of occupancy
rights, it is unclear why the wrongdoers’ second- or third-generation descendants—who
typically enjoy the ill-gotten territory in full (or at least adequate) knowledge of the moral
impropriety that produced their enjoyment—do not also count as complicit in the
wrongdoing (like the knowing recipients of stolen goods).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
from our family home and then succeed in raising your children and grand-
children in it (while my displaced family manages acceptably in the new
place to which we’ve fled), your family acquires an uncontested right to my
home. Indeed, even if my family doesn’t manage acceptably in our new loca-
tion, we retain at most a right to come back and share the house with your
family. This to me has less the look of a reasonable “balancing” of compet-
ing legitimate claims to the home than it does the look of a disturbingly
quick derivation of right from might. While Stilz assures us that her account
of territorial rights is not “a mere legitimation of the status quo,”47 it must
be allowed that a very strong conservative bias colors at least her treatment
of the “defeater” claims involved in cases of wrongful displacement.
Consider now Stilz’s account of wrongs of annexation. Because wrongful
annexation is a matter of violating a people’s right, according to Stilz, it fol-
lows that there is nothing obviously wrong with states annexing as much
territory as they please, provided that the territory in question is occupied
only by politically unorganized groups or individuals (or by [non-people]
subjects of some illegitimate state). While Stilz says that annexation must
still be “justified” even in such cases,48 it is very unclear how a legitimate
state could fail to be justified in such annexations—given the basic Kantian
(coercively enforceable) duty on each person to submit herself to legitimate
institutions for the administration of justice. But this appears to mean that
any legitimate state possesses a more-or-less permanent right to annex any
territory that is occupied only by persons who are not citizens in another
legitimate state—even by, say, that land’s harmless original occupants—and
this despite the fact that such persons may constitute no threat at all to
anyone in any legitimate state, at most “threatening” only those who freely
choose to share their lifestyle there.
Indeed, given that such persons do not enjoy “legal residence,” existing
legitimate states would appear to be entitled not merely to annex their land
but also to expel the land’s inhabitants. Stilz plainly wants to deny this,
affirming such inhabitants’ rights of occupancy; for she now characterizes
individuals’ occupancy rights as “preinstitutional” and uses as a central illus-
tration the case of the Navajo tribe, who “were not a politically organized
group.”49 But notice how she defines this preinstitutional occupancy right:
“the right of individuals to live in a certain area, and, together with others,
to authorize a legal institution to enforce rules regarding ownership, or to
engage in social practices defining their ownership.”50 Here, Stilz appears
47
Stilz (2011), 599–600 (n 42).
48
Stilz (2011), 598 (n 39).
49
Stilz (2013), 324 (n 1) 333.
50
Stilz (2011), 333.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
to begin parting company with Kant; for with mere “social practices” gov-
erning relations of ownership—rather than with a definite political/legal
institutional structure doing the job—the resulting property rights may
well lack the determinacy, stability, and enforceability required for them
to be (in Kant) “conclusive” moral rights. More important, though, those
individuals who are not engaged with others in collectively defining owner-
ship appear on this account to be denied preinstitutional occupancy rights,
making them morally vulnerable, if not to expulsion, then at least to subjec-
tion (through annexation of their land) by any legitimate state. But surely
even in such cases there is something wrong with simply taking or with
subjecting to outside control land that is being innocently and productively
used by individuals, at least where there is plenty of land left for others to
use as they please. The wrong in question here may involve neither wrongful
displacement of individuals (since such persons may be subjected to state
authority rather than displaced) nor wrongful annexation of a people (since
such persons lack a “people’s” political/legal organization). It is, instead, a
simple case of (what we can call) “wrongful subjection.”
As examples of such wrongful subjection, consider first an individual,
living in isolation from other persons (such as an idealized version of
Thoreau), who develops a conception of the good that is both firmly
opposed to membership in any political organization and revolves crucially
around his relationship to the particular land on which he resides. Consider,
next, a loose group of individuals (such as American frontier settlers, living
on widely separated bits of land) who mostly just mind their own business,
but who also together develop the same kind of ruggedly independent,
anti-political, territory-specific conception of the good. Finally, imagine a
socially well-integrated group—but one that lacks the formal, rule-centered
structure necessary for legitimate government and law (being, e.g., both
anti-democratic and “impressionistic” with respect to property and basic
rights)—that also has an anti-political and territory-specific conception of
the good (such as the Seminole “nation,” prior to its expulsion from Florida).
Now suppose that these individuals and groups are forcibly subjected by
others to (structurally just) institutions of government and law, making a
good life for them impossible.
Stilz appears to be committed to denying that these acts of forcible sub-
jection are wrongs (except, perhaps, in the case of the Seminole, depending
on how we understand their social practices regarding property). Her rea-
son, I suppose, must be that where people live outside of legitimate states
(or legitimating “social practices”), they live in a condition of injustice with
respect to those around them. If legitimate states subject (or expel?) such
persons, their doing so is defensible as an act required by justice. But if that
is indeed her answer, surely it misses the most important part of what
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
51
Stilz (2011), 334.
52
Stilz (2011), 581, 582.
53
Stilz (2011), 597.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
Bibliography
Altman, Andrew and Wellman, C. H. (2009). A Liberal Theory of International
Justice (Oxford: Oxford University Press).
54
Stilz (2011), 583.
55
Stilz (2011), 600.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
7
Can Reductive Individualists Allow
Defense Against Political Aggression?
Helen Frowe
INTRODUCTION
that seek to defend only political goods are necessarily disproportionate and
therefore always unjust. The Proliferation Problem holds that there is no
morally significant difference between states and some other collectives.
Some non-state collectives can embody whatever values we think warrant
lethal defense of the state (Rodin uses the example of the Cadbury chocolate
company, whose employees formed a distinctive community that was
destroyed when Cadbury was the subject of a hostile takeover by Kraft in
2010). So, even if we could show that it is proportionate for states to wage
defensive wars against threats to non-vital interests, we would lack grounds
for restricting this permission to states, which seems like a worrying result.
Rodin concludes that we should grant defensive rights to neither states
nor other collectives, and his conclusion thus results in a form of limited
pacifism. Unless a state is faced with genocidal aggression, Rodin thinks it
impermissible to resist for the state to forcefully resist invasion. It’s worth
noting, then, that Rodin is not conceiving of his argument as an objection to
reductive individualism. On the contrary: he thinks it’s true that states lack
a right to defend non-vital interests. I argue here that it’s not true, and that
reductivists aren’t committed to this claim.
In section 1, I outline the Conditional Force Argument in more detail. This
argument is underpinned by a particular account of our moral responsibility
for what I call “mediated harms.” In section 2, I argue that this particular
account of mediated harms is mistaken. Part of the problem with this
account is its reliance on a distinction between lesser interests and vital interests.
I argue in section 3 that this distinction also renders Rodin’s account of the
difference between political aggression and genocidal aggression incoherent.
In section 4, I defend a different account of our moral responsibility for
mediated harms. In section 5, I argue that the defense of aggregated lesser
interests can warrant endangering vital interests. In section 6, I consider the
Proliferation Problem, and argue that states (unlike Cadbury) protect our
right to form communities. It’s the right, and not communities, that matter
for permissible defense.
Imagine that a state wants to expand its territory by annexing land belong-
ing to another state. The aggressor state will use force only if the members
of the victim state resist the annexing. This kind of aggression—which will
become violent only if it is resisted—is what is sometimes called in the liter-
ature a bloodless invasion or, as I will call it, conditional force. Most accounts
of jus ad bellum hold that this kind of annexing of land constitutes a just cause
for defensive war on the part of the victim state. One plausible explanation
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
So, the annexing of land described above counts as an act of political aggres-
sion on this account.
Genocidal aggression is aggression that directly threatens vital interests—
that involves killing and otherwise harming the vital interests of members
of the victim state even if they do not resist. Such harming could be the end
of the aggression—it could, for example, be a war aimed at ethnic cleansing.
Or, it might be a means employed to some other end, such as access to
resources, but where no opportunity is offered for the victim state to capitu-
late. Rodin’s account somewhat stretches our ordinary usage of the term “gen-
ocidal” to include, for example, widespread raping, maiming, enslavement,
long-term displacement, and rending people stateless, but this extension
needn’t trouble us here.
Rodin argues that political aggression and genocidal aggression “have a
fundamentally different moral structure. Resistance to political aggression
will typically produce greater loss to vital interests among individuals within
the defended group, compared with engaging in no defence at all” (Rodin 2014:
87). Given this, resistance to political aggression is usually “straightfor-
wardly self-defeating and hence morally irrational” (Rodin 2014: 82). In
addition, since resistance involves risking the lives of other people for the
lesser interests of political goods, resistance is morally impermissible. If no
member of the polity may kill or endanger lives to defend her individual polit-
ical rights, then the polity as a whole may not kill or endanger lives to
defend those rights. Thus, it is only genocidal aggression that may be per-
missibly resisted, since here one endangers vital interests in order to protect
vital interests.
According to Rodin, the only way to explain why something like sover-
eignty warrants lethal protection is to attribute some kind of inherent value
to the state or political community (Rodin 2014: 69). This move is not
open to the reductive individualist, given her rejection of the idea that there
are collective values that are irreducible to their value for individuals. So, it
looks like the reductivist is committed to saying that sovereignty may not be
defended by means of war after all. War always involves inflicting lethal
harms and will therefore always be disproportionate when it is aimed at the
defense of the lesser interests of individual citizens.
tionate. And should such resistance be met violently—well, then the vital
interests of the citizens are directly threatened, and forceful resistance will
be a proportionate defense of those vital interests.
But Rodin argues that this reply fails because any calculation about the
proportionality of this non-violent resistance must take into account the harms
that the victim state foresees will be inflicted by the aggressor in response
(Rodin 2014: 82). Call these harms “mediated harms.” If the victim state pre-
dicts that its initial defense will cause the aggressor to escalate the situation,
that initial defense can be disproportionate. We can see this by thinking about
cases like Pinch:
Pinch: Bully wants to pinch Victim’s arm, which will hurt. Victim knows that if (and
only if) he tries to prevent this, Bully will become so enraged that he will kill five people.
It seems plain (to both me and Rodin, at least) that Victim may not try to
prevent Bully’s pinching his arm.2 Rodin grants that since Bully (and not
Victim) will inflict these mediated harms, they don’t weigh as heavily in the
proportionality calculation as harms that Victim will inflict himself. But the
fact that the harms will arise through someone else’s intervening agency
doesn’t make them irrelevant to what Victim ought to do. While mediated
harms are somewhat “discounted”—they don’t count as much for propor-
tionality as harm that one directly inflicts oneself—they are not discounted
to zero. They still matter for determining what one is permitted to do.
If this account of our responsibility for mediated harms is correct, the
reductivist must hold that the same reasoning applies at the level of war.
When the members of a state face a threat, they cannot judge whether war
or even non-violent resistance would be a proportionate response to that
threat unless they factor in how the aggressor will likely react. If they predict
that the aggressor state will react by waging a war that endangers the lives of
many people, they must take those foreseen harms into account.
Of course, the harms that the aggressor state will inflict are discounted in
the victim state’s proportionality calculation. But, Rodin argues, when it
comes to war, this discount is effectively canceled out because of the duty of
care that the members of the victim state owe their fellow citizens, which he
illustrates with the following case:
2
We might object that to prohibit Victim’s defending himself is to incentivize wrong-
doing. There are two reasons why I don’t think this shows us that Victim may defend
himself in Pinch. The first is simply that having any kind of proportionality constraint
makes it possible for a person to threaten you under circumstances in which it is imper-
missible for you to respond. Second, there’s no reason why Victim can’t factor predicted
long-term mediated harms into his proportionality calculation. If, in the long run, it
seems likely that acquiescing to a particular threat will incentivize wrongdoing that
causes more mediated harm overall than resisting the threat, it could be proportionate for
Victim to resist.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
House: A villain invades and occupies your home without justification. He lives in
your house and eats your food, forcing you and your family do all the work. He
makes it clear that he will not use force unless you resist. There is no end in sight.
You could tackle the invader, but if you do so, it is likely that one of your children
would be killed, either as a side effect of the struggle, or as a punishment for your
resistance. (Rodin 2014: 84–5)
Rodin argues that most people would think it impermissible to tackle the
villain, because “the value of a child’s life – especially since this is your child
and you owe a profound duty of care to him” outweighs the good of evict-
ing the occupier (Rodin 2014: 85). He thinks that a similar duty obtains
when members of a state are considering going to war. The people whom
they will be endangering through their resistance are “people who are
bound to [them] by relationships of loyalty, community and kinship.”
They “may be our comrades in arms, our family members, our neighbours”
(Rodin 2014: 83). So, any discount that arises from the fact that it is the
aggressor state that will be endangering these people is undone by the fact
that the members of the victim state owe a duty of care to the people who
are endangered.
The members of the victim state must, therefore, proceed as if they them-
selves are harming their co-citizens as a side effect of their self-defense. Most
people think that one may not, for example, kill a bystander as a side effect
of defending oneself against a broken leg. Similarly, Rodin argues that one
may not endanger one’s fellow citizens’ vital interests for the sake of defend-
ing something of lesser interest, which for Rodin includes land, resources,
and political independence. Exposing people to such harms is dispropor-
tionate compared to the goods threatened by purely political aggression.
Reductivism is, therefore, unable to allow that resistance against political
aggression can be just.
While Rodin’s account of meditated harms seems to get the right result in
Pinch, it’s less plausible in other cases. Take Rape:
Rape: Angry Rapist tries to rape Alice. If (and only if ) Alice fends him off, he will be
so angry that he will go and rape two other women. If (and only if ) Alice kills Angry
Rapist, Angry Rapist’s friend will go and rape two other women.
It seems to me very plausible to think that Alice may defend herself against
Angry Rapist even if she foresees that her defense will trigger the rape of two
other people. And, yet, the harm she predicts will arise from her defense is
twice that which will befall Alice herself. Even if we discount the foreseen
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
harm by fifty percent, in virtue of the fact that it is mediated harm, Alice
would still be triggering the equivalent of one rape as a side effect of defend-
ing herself against rape.
Rodin’s view entails that Alice must proceed as if she would be inflicting
the discounted harm herself—as if she would be inflicting the rape upon the
other woman as a side effect of her defense. But most accounts of permissi-
ble defense do not allow someone to inflict as a side effect a harm that is
equal to the harm she is seeking to avoid. For example, I may not divert a
runaway trolley from where it will kill me down a sidetrack to where it will
kill some other innocent person. Rather, I would be permitted to divert the
trolley only if the person on the sidetrack would suffer a significantly lesser
harm than the harm facing me. If this is right, even if we discount the medi-
ated harm by fifty percent, Rodin’s account still prohibits Alice’s defending
herself in Rape.
This seems like the wrong result to me. Rodin might reply that mediated
harms can be more heavily discounted than I have suggested here. As I will
argue in section 4, I think they are much more heavily discounted. But this
isn’t going to help Rodin’s account, because this view requires that however
much we discount the mediated harms, the duty of care that we owe to our
co-citizens cancels out this discount. This means that Alice may not fend off
Angry Rapist when doing so will cause him to rape or seriously assault just
one other woman if that the other woman is a co-citizen.
Of course, Rape illustrates the defense of a vital interest—the right not to
be raped. We might think that we should therefore treat Rape differently to
Pinch, which involves the defense of a lesser interest—the right not to be
pinched. We might say that while mediated harms can make defense of lesser
interests disproportionate, they can’t make the defense of vital interests dis-
proportionate. We may always defend our vital interests even if we foresee
mediated harms to other innocent people’s vital interests.
But this response is pretty ad hoc: it’s not clear why our account of
responsibility for mediated harms should be sensitive to this distinction
between lesser and vital interests. As Rodin says, “[d]efensive action is
impermissible when it foreseeably produces harmful effects that are dispro-
portionate to the good one is seeking to achieve” (Rodin 2014: 77). This claim
cannot plausibly be restricted to the defense of lesser interests—it’s simply
true of defensive action in general, which includes the defense of vital inter-
ests. If mediated harms count among our harmful effects—as Rodin and
I think they do—they must be able to make defensive action disproportion-
ate even when one is defending vital interests.
Moreover, it’s not clear that this response is even open to a proponent of
the Conditional Force Argument. Given, for example, Rodin’s claim that
“[t]he morality of resisting the direct threat component of a conditional
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
This line of argument also reveals two difficulties with the distinction between
political aggression and genocidal aggression that corresponds to the distinc-
tion between threats to lesser interests and threats to vital interests. The first
concerns Rodin’s claim that his account allows that “coordinated, forceful
defence against genocidal aggression can be morally justified in reductivist
terms” (Rodin 2014: 83). We’ve just seen how Rodin’s account makes it
impermissible for Alice to defend her vital interests if, in doing so, she will
expose others to similar harms. As we saw, it could be the case that, even
discounted, the mediated harms are similar in weight to the harms facing
Alice, and thus she would not be permitted to defend herself while foresee-
ing those harms to others.
If so, it must also be true—if we are reductivists—that states may not
defend their vital interests when doing so exposes others to similar harms (as
defensive war inevitably does) even once we apply the discount. For exam-
ple, a state with a small population facing a genocidal threat may foresee
that resisting will result in a war that will cause lethal and other very serious
harms to a large number of people: perhaps more people than whose lives
are threatened by the genocide. If resistance risks the vital interests of twice
as many people as are threatened by the genocide, even discounting these
mediated harms by fifty percent will not make resistance permissible, just as
it was not permissible for Alice to resist if she foresees that the equivalent of
one other rape will be inflicted on another innocent person. In other words,
Rodin’s account of mediated harms will sometimes prohibit resisting even
genocidal aggression.
The second problem is that there will be cases in which Rodin’s account sim-
ply fails to classify an act of aggression as either political or genocidal, and
thereby fails to tell us whether such aggression may be resisted. Consider Threat:
Threat: An aggressive state is demanding the execution of all 1000 members of a
small ethnic minority group living in the victim state. If the victim state does not
kill these citizens or permit a delegation from the aggressor state to do so, the
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
aggressor will invade. During the invasion, 100 members of the minority group will
be killed, along with 1100 defending combatants and innocent non-combatants in
the victim state.
So, a total of 1200 innocent citizens will suffer lethal harm if the victim state
decides not to capitulate to the aggressor’s demands to execute the 1000. Must
the victim state capitulate?
What Rodin is going to say about Threat will depend on how he is identify-
ing the relevant group, harms to which we must factor into our proportion-
ality calculation. His definitions of political aggression and genocidal
aggression are tied to thoughts about the harms that will befall the group being
defended—he speaks of “those very people whom the defensive action is
intended to defend ” and “the defended group” (Rodin 2014: 87). The ques-
tion, then, is whether these references to people and groups are meant to
pick out all members of the victim state that is being conditionally threat-
ened, or just the sub-group that is being directly threatened (in this case
the members of the ethnic minority).
If the “group being defended” is meant to be synonymous with “all mem-
bers of the victim state,” that will give us the result that the aggression in
Threat is political aggression by Rodin’s lights. Fewer vital interests will be
harmed overall within this group if the victim state capitulates and executes
the members of the ethnic minority. Resistance against this aggression will
thus be impermissible. This result seems to me wrong in both its denial that
the threat is genocidal, and in its seeming pronouncement that the victim
state ought to execute the members of the ethnic minority group rather
than defend them.
So, I should think that Rodin will want to go the other way, and argue that
references to “the group being defended” are not meant to be synonymous
with “all members of the victim state,” but can rather refer to a sub-group
within that state, such as the ethnic minority group I have described. He
tells us that an aggression is political, and should not be resisted, if resisting
it causes more harm to the vital interests of “those very people whom the defen-
sive action is intended to defend.” Since in my example, resisting will be less
harmful to the vital interests of the minority group whom the action is
intended to defend, Rodin can count the threat as a genocidal aggression.
But this result isn’t just semantic in his view—it’s normative. It tells us
that the invasion may be resisted, because it’s genocidal. And that seems
right. But this is a case in which more harm will befall the vital interests of
innocent people overall if the invasion is resisted, which was meant to be the
definition of political aggression and part of the explanation of why resist-
ing political aggression is wrong: resisting in such circumstances is, accord-
ing to Rodin, morally irrational. So we have two problems. One is that this
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
I think that our judgments about how much harm Victim ought to bear to
rescue people from harm can usefully inform our judgments about how
much cost Victim must bear in a mediated harm case rather than endanger
some other innocent person.
The connection between the two sorts of case might be explained by
Victor Tadros’s suggestion that both the restrictions on harming people as a
means for the greater good and the restrictions on requiring people to bear
costs to rescue other people “have as their source our status as ends. Because
we are not available as tools for the sake of each other, we cannot be harmed
as a means to a greater good and duty does not require us to act for the greater
good” (Tadros 2011: 250). If Alice were not permitted to defend herself against
Angry Rapist, we would effectively be requiring Alice to treat herself as a
means, permitting Angry Rapist to harm her for the sake of others. We can
sometimes be so required, but there are limits to how much cost we can be
required to bear to such ends, just as (and for the same reason as) there are
limits on the costs we must bear to rescue others.
3
Even if we accept Rodin’s duty of care argument, which could explain why we need
not care too much about harms to non-combatants in the aggressor state, this will not
explain why the vital interests of other co-citizens can be ignored in our calculations
about whether resistance is proportionate.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
This result is more permissive than Rodin’s view. Rodin denies that there is
this sort of ratio between foreseen mediated harms and the defended good,
holding simply that whenever resistance will harm a greater number of vital
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
interests among members of the victim state, the aggression is political and
the state must capitulate. This is because of his view that the discounting of
harms caused by an aggressor state in war is neutralized by the duties of care
that one owes to one’s co-citizens. One must proceed as if one were directly
inflicting the harms upon those co-citizens.
Let’s assume for a moment that Rodin is correct that there is a duty of care
that requires us to treat our co-citizens with special care. If there is such a
duty, it will apply to all co-citizens. This means that one will have a duty of
care not only to those whom one foresees will be harmed by the wrongdoing
of others, but also to those whom one can protect by resisting an invasion.
So, we will need to think about how this duty plays out in a case like Threat,
where one must choose between failing to protect a smaller number of citi-
zens from harm and causing mediated harms to a larger number of citizens.
Rodin argues that one may not intentionally kill an innocent person to
save the life of another person to whom one owes a duty of care. I may not
throw someone else’s child in front of a trolley to save my own child’s life.
But the harms to the larger number of citizens in Threat—and more gener-
ally in defensive war—are not intentional, but merely foreseen mediated
harms. We might think that this makes them easier to justify. However,
Rodin’s House case that I described in section 1 is meant to illustrate that
even when one merely foresees a mediated harm to another person’s vital
interest, one cannot bring that harm about when one owes a duty of care to
the person who will be harmed.
But this case seems to me unpersuasive, not least because it should count
as a case of genocidal aggression on Rodin’s account, and he claims that
genocidal aggression may be resisted. Rodin lists enslavement among the
infringements of liberty sufficiently serious to count as genocide. One would
also think that indefinite imprisonment counts as a sufficiently serious
infringement. To preserve the analogy between House and war, it must be
the case that in addition to being made to “do all the work,” you and your
family cannot leave the house (and even if you can, Rodin counts indefinite
displacement from one’s home as a genocidal threat when it happens at the
national level) (Rodin 2014: 79; Rodin 2014: 75).
Given this, House is a case in which more people will suffer setbacks to
their vital interests if the occupation is not resisted. Even if it is certain that
one child will die in the resistance, not resisting entails that the whole family
will be indefinitely imprisoned and enslaved. It is, by Rodin’s own lights, a
case of genocidal aggression.
So, may you resist? If Rodin’s claims about genocidal aggression are true,
then he must change his mind and say that you may—given the greater set-
back to vital interests overall, this is not a case in which it would be morally
irrational or impermissible to resist. More harm to vital interests befalls the
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
very group of people being defended if the occupation is not resisted, and
all members of the family are owed a duty of care.
If Rodin sticks to his guns and says that you may not resist, this implies
a ranking of interests even within the category of vital interests, according
to which it is always impermissible to expose someone to the top-ranked
harm—death—even if a greater number of other vital interests (the rights
not to be enslaved or imprisoned) are thereby harmed. But as I pointed out
above, a central premise of the Conditional Force Argument is that one
cannot stipulate the proportionality (and impermissibility) of resisting a
direct threat without taking into account the foreseen harms. So, Rodin
cannot stipulate that death always outranks harm to other vital interests,
such that it is always disproportionate to foreseeably bring about some-
one’s death while defending against a threat to less serious (but still vital)
interests. Such a stipulation would anyway be implausible. Foreseeably
causing one person’s death as a mediated harm does not seem dispropor-
tionate to freeing a larger number of people from indefinite imprisonment
and servitude.
I think it is possible that familial ties and ties of friendship can sometimes
make a difference to what we are allowed to do, although it’s not clear to me
that this is best explained as owing a duty of care to these people. But if I do
owe a duty of care as generally as Rodin supposes—essentially, to all those
people who participate in my state—it seems to me unlikely that I owe it to
them in virtue of the fact that they participate in my state. Such a duty must
rather be grounded in more general properties shared by all persons. That
this is simply a general duty owed to all persons explains why, for example,
I may not drive more recklessly in a popular tourist area, or in a foreign
country, than I drive at home.4 Such reckless endangerment could not be
justified by appeal to the absence of a duty of care, or by the claim that any
such duty is less stringent when it comes to foreign people. Nor are my
duties to rescue less stringent when I’m abroad, where all the imperiled
people are foreign. Nor does the fact that one of two drowning people hap-
pens to be British give me a reason to save her rather than the other person.
If there is a duty of care that explains why I must care about all these harms,
I should think that duty applies equally to foreigners and co-citizens. If I
can divert a runaway trolley away from where it will harm five to either:
(a) where I foresee that it will harm Local Person, or (b) to where it will
harm Foreign Person, I don’t think a duty of care requires me to choose (b)
and harm Foreign Person rather than Local Person.
4
Rodin does say that the duty of care is less important when it comes to intentionally
harming. But reckless endangerment is not intentional harming, and the tort law duty of
care is generally thought to cover just such endangerment.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
This brings us to the crux of defense against political aggression: the jeopardiz-
ing of some people’s vital interests to defend others against threats to their
lesser interests. In making his binary division between vital and lesser inter-
ests, Rodin assumes that no number of threats to lesser interests could out-
weigh the harming of a vital interest. I think we should reject the division
between lesser and vital interests. There are some interests that will never
warrant lethal defense—there is probably no number of scratched fingers
that could make it permissible for me to avert those scratches by endangering
an innocent person’s life. But, as I will argue, some of the things that Rodin
counts as lesser interests are going to warrant lethal defense when they are
aggregated across a sufficient number of people, and also warrant triggering
mediated harms to the vital interests of innocent people. In other words,
their defense will justify both intentional lethal harms to combatants and
foreseen but unintended lethal harms to non-combatants.
Here’s an example of how so-called lesser interests can aggregate to make
lethal force proportionate. Imagine that you are going to break my arm. I
don’t think that I may kill you to stop you from breaking my arm. But if
you’re going to break lots of people’s arms, it seems to me that there will
come a point at which these harms can aggregate to make it proportionate
to kill you to prevent them (of course, other conditions such as necessity
would also need to be satisfied for killing you to be overall permissible). This
aggregation can be both interpersonal and intrapersonal. If every time my
arm heals, you break it again, I think lethal force could become proportion-
ate to stop you from continuing to break my arm. I think this is also true
when it comes to my political rights. As Rodin says, I may not kill you to
stop you from preventing me from voting. But if you’re going to stop every
member of my ethnic group from voting in an election where our impor-
tant political interests are at stake, it might well be proportionate for us to
kill you so that we can vote. And if killing you is necessary to avert this
wrongdoing, it might well be all-things-considered permissible to kill you.
It might seem as if invoking aggregation is an illegitimate move for the
reductive individualist to make, since it seems to go beyond what can be
justified in terms of individual rights. But that would be to misunderstand
the reductive individualist view. The reductivist part of the account requires
us to understand the morality of war in terms of the morality of ordinary
life. Aggregation is a familiar feature of ordinary life, and so to think it rel-
evant to war is not inconsistent with reductivism. The individualist part of
the account holds that individuals are the source of moral value and the
proper focus of moral prescription and evaluation. But this too is compati-
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
ble with caring about the number of individuals who are threatened with a
violation of their rights. All that matters for consistency with individualism
is that, ultimately, it is the individuals themselves with whom we are con-
cerned, and not with some irreducible relationship between those individu-
als. The claim that I am defending here is that a conditional invasion threatens
important rights of individuals, not of collectives.
Numbers are significant when it comes to political rights not merely for
straightforward aggregative reasons, but because the increasing numbers
can produce a step-change in what threatens each individual.5 If you dam-
age my car and prevent me from voting, this won’t really make any differ-
ence to the sort of country I live in. But imagine that measures are taken
before a UK election to ensure that anyone living outside of London is
unable to vote. This would do much more than frustrate each individual
non-Londoner’s right to vote—it would not be simply a case of iterating the
same harm of vote prevention across many individuals. Rather, the increas-
ing numbers would effect a change in the sort of harm being perpetrated
against each of those individuals, because it would change the UK from a
legitimate democratic state to an illegitimate dictatorship. States coercively
impose legal frameworks upon their citizens. A government that is elected
only by Londoners, but then attempts to rule over the whole of the UK, will
be imposing this framework illegally, unjustly coercing individuals to adhere
to it. This is a much more serious wronging of each individual citizen than
the wrong of preventing any one individual from voting, and it is this sort
of wronging that is threatened by a political invasion. An aggressor state
that seeks, for example, to replace a democratically elected government, and
threatens to meet resistance with violence, significantly worsens the condi-
tions of the citizens’ lives irrespective of what else the aggressor does.
I’ve argued that Rodin is wrong about the impermissibility of resisting con-
ditional threats. I’ve done this in a way that doesn’t depend upon attributing
irreducible moral value to the state. But reductivists also need to say some-
thing about the Proliferation Problem. The challenge is not exactly Rodin’s
suggestion that other non-state groups can manifest the inherent value found
in the state. Rather, it’s that if (as I believe) there’s nothing uniquely and/or
inherently valuable about states, we want to know whether other groups of
individuals also have rights to employ lethal defense against the equivalent
of “political” threats to their integrity. To use Rodin’s example, we want to
5
Thanks to Jimmy Lenman for helpful discussion of this point.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
Thus far, we’ve been talking about threats to interests that are, uncontrover-
sially, protected by rights. We have rights not to be killed, be raped, be
imprisoned, be prevented from voting, have our arms broken, and so on.
We’ve assumed throughout that the victims of threats to their interests are
innocent people who have done nothing to forfeit their rights that protect
those interests. Even if Rodin and I disagree about whether one may violently
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
defend one’s right to vote, we presumably agree that one has such a right that
is (ordinarily) violated if one is prevented from voting.
Are any rights violated by the Kraft takeover of Cadbury? I do not think
that they are. Rodin describes Cadbury as being built along Quaker values
of dignity and respect, with a focus on providing a holistic approach to
employee welfare, including the provision of high-quality housing, subsidized
recreational activities, and funding service to the community. We can see
that the employees certainly had an interest in the continued provision of
these goods and the continued existence of this community. But not all of our
interests are protected by rights.
I’ve argued that, at the level of the state, successful political aggression
backed by conditional force effects a significant change in the sort of soci-
ety that citizens live in. It worsens their lives by depriving them of impor-
tant goods such as their freedom elect their government, to protest against
unjust treatment and so on. These goods are things to which people have
rights, and thus political aggression at the level of the state violates those
rights.
But this doesn’t seem true in the case of a takeover by Kraft. This is, in
part, because I don’t think the employees had rights to the continued provi-
sion of the goods provided by Cadbury. While we have a right to try to form
communities, rooted in our broader and very important right of free asso-
ciation, we do not obviously have rights to the continued existence of our
communities. If, for example, I start a War Ethics club, this might be an
enterprise that is very valuable to me, that benefits me in various ways and
so on. But if the other members of the War Ethics club foolishly decide that
they have better things to do with their time, I have no moral power at all
to try to coerce them into staying, even if their departure brings about the
collapse of my valuable community. What I have is, as I said, a right to free
association that includes a right to try to form communities and other rela-
tionships, but not a right to the continuation of any particular community
or relationship.
It’s this right that is protected by the state, and that is not protected by a
group such as Cadbury nor threatened by a group such as Kraft. Notice that
it was perfectly possible for those ousted from Cadbury by Kraft to go and
set up another community, run along Quaker lines, elsewhere. Kraft might
be able to destroy the employees’ existing community, but it has no power
to prevent them from establishing other or replacement communities.
Importantly, it’s the state that provides the protected space in which those
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
employees can try to form a new association. I think this feature of a state
helps explain why states (and some other entities) can warrant lethal defense
while an entity like Cadbury cannot. The state protects not our particular
communities, but our right to form those communities.
We shouldn’t confuse this with the claim that the state is valuable because
it protects many communities, whereas a particular community like Cadbury
does not. Rodin considers and rejects this explanation of why states warrant
defense: “This proposal suffers from the problem [that if ] the state possesses
value because of the communities that exist within its territory, then we
would expect that states with rich ecosystems of local communities would
enjoy greater defensive rights than states with sparser communal activity.
They do not” (Rodin 2014: 74). Rodin is addressing, here, the idea that the
source of defensive rights lies in the value of the communities that are pro-
tected (hence, his claim that more communities ought to equal greater defensive
rights). But my claim is that the real value lies in the state’s protecting its
members’ rights to free association. Both the communities and the right are
valuable, but it’s the right that plays a role in establishing proportionate
defense.
This way of understanding the claim answers Rodin’s concern that states
with less communal activity will end up with less stringent defensive rights.
On this account, it’s not the communities per se that are defended, but the
right to form them. States that protect this right, along with its citizens’ other
important rights, are equally valuable in this respect, even if their members
choose to exercise their associative rights to different degrees.
There’s perhaps a simpler variant of Rodin’s reply that might be leveled at
my view—namely, that it implies that more populous states have more
stringent rights of defense than less populous states. After all, if the numbers
matter, then it seems that it will be proportionate for some states to do more
than others in the face of aggression, since they have a greater number of
members whose rights are threatened by aggression.
I think this is correct, but that this is not an objection to reductive individual-
ism. Rather, this seems to me an appealing account of defensive rights.
More populous states protect the interests of larger numbers of people, and
it’s therefore quite plausible to think that they may do more in defense of
those people. When the good protected is “larger” in this sense, it’s propor-
tionate to do more to protect it. There could be reasons why international
law ought not to dictate that some states may use only less harmful defense:
such a law would invite conditional aggression of a particularly violent sort
towards those states. But it may nonetheless be true as a matter of morality
that a state with a small population may not do as much to defend its mem-
bers’ political rights as a more populous state. But this is compatible with
thinking, as I have argued, that political aggression of the sort under discus-
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
sion here threatens serious rights, and that these rights are serious enough
that it will be proportionate for most (and probably all) states to use mili-
tary force to protect them.
11. CONCLUSION
Bibliography
Buchanan, Allen (1998). “Secession and Nationalism,” Goodin, R. and Pettit, P.
(eds), A Companion to Contemporary Political Philosophy (Blackwell).
Fabre, Cécile (2009). “Guns, Food and Liability to Attack in War,” Ethics, Vol. 120,
No. 1, 36–63.
Frowe, Helen (2014). Defensive Killing (Oxford: Oxford University Press).
Kamm, Frances M. (2014). “Self-Defence, Resistance and Suicide: The Taliban
Women,” in Helen Frowe and Gerald Lang (eds), How We Fight: Ethics in War
(Oxford: Oxford University Press).
Kutz, Christopher (2005). “The Difference Uniforms Make: Collective Violence in
Criminal Law and War,” Philosophy and Public Affairs, Vol. 33, No. 2, 148–80.
Lazar, Seth (2014). “National-Defence, Self-Defence and the Problem of Political
Aggression,” in Cecile Fabre and Seth Lazar (eds), The Morality of Defensive War
(Oxford: Oxford University Press).
McMahan, Jeff (2009). Killing in War (Oxford: Oxford University Press).
Rodin, David (2014). “The Myth of National Self-Defence,” Cécile Fabre and Seth
Lazar (eds), The Morality of National Defence (Oxford: Oxford University Press).
Tadros, Victor (2011). The Ends of Harm: The Moral Foundations of Criminal Law
(Oxford: Oxford University Press).
Tadros, Victor (forthcoming), “Unjust Wars Worth Fighting For,” Journal of Practical
Ethics.
Walzer, Michael (1974). Just and Unjust Wars (New York: Basic Books).
Zohar, Noam (1993). “Collective War and Individualistic Ethics: Against the
Conscription of Self-Defense,” Political Theory, Vol. 21, No. 4, pp. 606–22.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
8
Elbow Room for Rights
Eric Mack
1. INTRODUCTION
Robert Nozick’s first sentence in Anarchy, State and Utopia is the ringing
affirmation that “Individuals have rights, and there are things that no per-
son or group may do to them (without violating their rights).” Nozick imme-
diately adds that these rights may be even more restrictive than one might
anticipate. “So strong and far-reaching are these rights that they raise the
question of what, if anything the state and its officials may do. How much
room do individual rights leave for the state?” (1974, p. ix).
In this chapter, I address questions about whether core rights affirmed by
Nozick, viz., persons’ rights over themselves and over their legitimately acquired
extra-personal holdings, are considerably more deeply restrictive than Nozick
himself anticipates when he wonders whether these rights would morally hog-
tie aspiring statesmen and state officials. Indeed, I address questions about
whether those rights are so constraining that they deeply restrict their own
exercise. These questions have been forcefully pressed in two important critical
discussions of rights-oriented libertarian thought, Peter Railton’s “Locke,
Stock, and Peril: Natural Property Rights, Pollution, and Risk” (1985) and
David Sobel’s “Backing Away from Libertarian Self- Ownership” (2012).
Both of these papers maintain that the rights of self-ownership and of
private property in extra-personal holdings—which for convenience I will
refer to as the “libertarian rights” or the “Lockean rights”1—are much more
restrictive than their advocates generally recognize. The key thought is that
libertarian rights demand that all agents desist from even the most minor
unprovoked and non-consensual physical intrusions upon right-holders; in
1
Although I employ these terms to refer to rights of self-ownership and individual
entitlements to extra- personal objects, I do not mean to say that these are the only moral
rights libertarians should affirm or that self-ownership is the only natural moral right.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
addition, these rights at least seem to demand that all agents desist from
imposing on others even modest risks of such intrusions. Yet, compliance
with these demands requires that individuals forego all or almost all of the
ordinary exercises of rights by which individuals maintain their lives and
achieve their ends. As legal theorist Richard Epstein puts it, “If any smell,
noise, or discharge counted as a nuisance [and, thus, a rights infringement],
no one could barbecue in the backyard, talk on his front patio, or farm”
(2009, pp. 15–16).
The distinctive complaint that Railton and Sobel bring against stringent
libertarian rights is that such rights morally hog-tie their possessors. The
purported difficulty is that these rights systematically make their exercise
morally impermissible. Railton signs on to this complaint when he says,
I will argue that when one attempts to apply such [Lockean rights] theories to moral
questions about pollution, they present a different face, one set so firmly against
laissez faire – or laissez polluter – as to countenance serious restriction of what
Lockeans have traditionally taken to be the proper sphere of individual freedom.
(1985, p. 89)
Sobel signs on to this complaint when he says,
The ubiquity of difficult to avoid, minor infringements on other people’s bodies makes
the simple argument [from libertarian rights to standard libertarian conclusions]
unattractive. Strong moral constraints against all such infringements would make
too many things impermissible. The thought that, quite generally, my self-ownership
creates very powerful moral constraints on any and all involuntary infringements on
my body would unacceptably interfere with your liberty as Nozick saw. (2012, p. 35)
Could the philosophical theory named for liberty actually turn out to be unacceptably
restrictive of our liberty? (2012, p. 37)
The hog-tying complaint is not that the domains of some individuals may be
too small; i.e., that they may include too little in the way of personal or extra-
personal resources. Rather, the complaint is that much of whatever is within
one’s domain in the sense that others may not destroy or seize or control it
without one’s permission will, nevertheless, not be within one’s sovereign
domain in the sense that one’s chosen use of that material is morally allowed
and protected. Note that this complaint can be directed against any doctrine
that broadly ascribes rights to individuals over persons or extra-personal
objects that others are morally forbidden to infringe. For example, if rights
to extra-personal objects are ascribed to individuals on the basis of a strongly
egalitarian-leaning end-state principle and these rights are taken to be
infringed by any unconsented to physical intrusion, those rights will be subject
to the hog-tying complaint.
In this chapter, I focus on what the friend of libertarian rights or any other
advocate of similarly potent rights should say about “minor intrusions,” i.e.,
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
2
On pure risks, see Railton, p. 95.
3
The now common “property rule” and “liability rule” terminology is introduced in
Calabresi and Melamed.
4
Sobel (p. 38) says that Nozick’s “main response [to the apparent impermissibility of
minor intrusions] is to claim that our property rights do not create boundaries that it is
generally impermissible to cross. Rather, others may permissibly cross our boundaries
provided that they adequately compensate us for doing so . . .”
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
5
I have previously labeled this “the anti-paralysis postulate.” See Mack (2012), pp. 112–14.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
over extra-personal objects. For this reason, the abstract right of property—a
right to make things one’s own and to exercise discretional control over
what one has made one’s own—is also a crucial and salient dimension of
the proper articulation of the libertarian organizing principle.6
Is an individual who needs a kidney disallowed from living her own life
in her own chosen way by another individual declining to supply her with
his spare kidney or by his evading her attempts to extract that needed kidney?7
Is an individual who desires to engage in certain sexual acts precluded from
living his own life in his own chosen way by the sole object of his sexual
desires declining to participate in those acts or by her evasion of his advances?
The organizing principle for Lockean rights answers these questions in the
negative. Agents who are not enabled to live as they choose because another
party declines to supply one of his kidneys or declines to participate in
desired sexual interactions are not thereby made unable—or precluded from—
living their own lives in their own chosen ways. “Failing to help another
cannot be construed as interfering with his right to use himself as he
wishes . . .” (Cohen 1995, p. 215). Absent this understanding, ordinary exer-
cises of rights by one party will regularly also count as ordinary violations of
other parties’ rights and the moral claims of individuals to live their own
lives in their own chosen ways will systematically conflict. Any resolution of
that conflict would require the demotion of rights into moral commodities
that are to be traded off against one another.
I need to emphasize that it is not the role of armchair philosophy—even
natural rights philosophy—to discover and disclose the precise contours of
persons’ nitty-gritty rights. Those precise contours do not exist out there in
the nature of things or as theorems that are deducible from Lockean axi-
oms. So, it is not the business of a Lockean theory of rights to determine
whether or not the owner/operator of a well-established water mill has a
right against individuals living upstream that they not significantly dimin-
ish the flow of water that turns his mill. It is not the business of this or any
other philosophical theory to determine exactly how loud the noise has to
be that emanates from A’s property in order for B to have a right to enjoin
A’s drum-playing.
The relatively concrete rights that are reasonably ascribed to individuals
in a given society—e.g., the right not to be subjected to noise over a certain
decibel level—provide a structure of reasonably expected liberties and
immunities that facilitate peaceful coexistence and voluntary cooperative
6
A right against deceptive manipulation would be a third element within the abstract
framework of libertarian rights. For deceptive manipulation is yet another way in which
one can be precluded from living one’s own life in one’s own chosen way.
7
I thank Sobel for pressing this question.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
8
This sort of layered articulation of rights is suggested by Loren Lomasky’s distinction
between basic rights, moral rights, and legal rights (Lomasky 1987) and Gerald Gaus’
distinction between abstract moral principles and more concrete moral rules (Gaus 2012).
9
See Mack (2010), p. 63 and Gaus (2012), pp. 391–417.
10
See Epstein (1979) and (2009).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
capture much of what makes those rights attractive. For central to the
appeal of rights is that they provide individuals with moral protection for
their choices with respect to a range of alternatives. If one has rights over
oneself and over one’s holdings, one may dispose of oneself or one’s holdings
as one sees fit; one may do as one wills—subject, of course, to the standard
requirement that one is not violating others’ rights. As Locke held, each
person’s fundamental claim against others is to “. . . a Liberty to dispose, and
order as he lists, his Person, Actions, Possessions, and his whole Property,
within the Allowance of those Laws under which he is, and therein not to
be subject to the arbitrary Will of another, but freely follow his own” (1689,
Second Treatise, §57).
All the standard vocabulary of individual sovereignty, moral inviolability,
moral space, individual spheres of authority, and domains of personal free-
dom reflect the really basic idea that rights are about the protection of
choice. Persons may be harmed in certain ways when their rights are vio-
lated and that harm may be counterbalanced by simultaneous or subse-
quent due compensation; but such compensation does not counterbalance
the wrong that is done to the right-holder by depriving him of the choice of
whether he will be subject to that counterbalanced harm. The slate is not
wiped clean by due compensation because compensation deals with the
harm and not the wrong.
What if there is some extra compensation for being deprived of the choice
about whether one will be subjected to the counterbalanced harm? This
question suggests that having choice in the matter is itself just another inter-
est of the right-holder that is not to be set back without due compensation.
But even if we say that having choice in the matter is an interest of the right-
holder, it remains a very special sort of interest. For the claim to having the
choice about whether to be subjected to counterbalanced harm cannot be
honored by depriving the claimant of that choice and throwing in a bit of
extra payment. David Schmidtz rightly emphasizes that at the core of a
property right—indeed, at the core of any moral claim protected by a prop-
erty rule—is “a right to say no to proposed terms of exchange” (2010, p. 79).
One’s right to say no is not honored by others infringing upon that right
while providing one with the payment one would have accepted if one had
not said no. One’s right to say no is not honored if the treatment one objects
to is imposed upon one—even if it is true that, had one deliberated well,
one would have waived one’s right against the treatment in exchange for the
payment that is now offered as compensation.
A general liability rule attenuation of rights does not comport with
Nozickian invocations of moral inviolability (1974, pp. 31–2). One salient
dimension of belief in the moral inviolability of persons is subscription to
principled anti-paternalism. According to principled anti-paternalism, at
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
least some interventions that aim at the recipient’s good and on balance
do advance the recipient’s good are still impermissible infringements of
the recipient’s rights. However, liability rule attenuation undermines prin-
cipled anti-paternalism. For, given that attenuation, if an intervention
bestows an over-balancing good on its recipient, there will be no infringe-
ment of rights.
Nozick seeks to elucidate moral inviolability by contrasting beings who
possess inviolability with tools (1974, p. 31). But, how much less of a tool is
one if one has rights that are protected by liability rules rather than having
no rights? Suppose that there were a moral norm according to which anyone
who uses a particular unowned hammer is morally required to repair any
damage done to it or even to shine the hammer up a bit. It does not seem
that such a norm would make the hammer any less of a tool than it would
be without protection by that liability norm. What more is present when
everyone else is required to duly compensate you—to repair any damage to
you or even shine you up a bit—after making use of you? The only thing
more that is present is that you—unlike the hammer—have a claim to that
compensation. Not only is there a price for using you; the payment of that
price is owed to you. So, your having a right against being used in certain
ways that is protected by a liability rule is a little bit more than there being
a moral norm that says that whoever makes use of you in an intrusive way
must repair you or even shine you up a bit. Nevertheless, it is very difficult
to see that this little bit more makes you significantly less of a tool than that
unowned hammer.
In Anarchy, State and Utopia, Nozick devotes several trenchant pages to
his rejection of H. L. A. Hart’s principle of fairness (1974, pp. 90–5).
According to this principle, individuals who have taken on certain bur-
dens in the course of some benefit-conferring activity have a right against
all beneficiaries of that activity that they share in those burdens. So, if you
have benefited from all your neighbors taking time to tell funny stories on
the neighborhood public address system, you are bound yourself to spend
some time telling funny stories (or the like). Nozick rejects this saying,
“One cannot whatever one’s purposes, just act so as to give people benefits
and then demand (or seize) payments” (1974, p. 95). Yet, if one accepts
the liability rule attenuation of rights, why shouldn’t this be allowed?
Why isn’t the benefit already conferred by those who now demand (or
seize) payment simply advance compensation for the boundary-crossing
they subsequently engage in? If one may seize as long as one makes due
compensation afterwards, why may one not make due compensation for
a future seizure?
As one would expect, a general liability rule attenuation of rights renders
many actions morally permissible that the friend of libertarian rights plausibly
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
takes to be violations of rights.11 Ceteris paribus all forced exchanges that leave
their subjects at least as well off as they would be were they not subject to
the forced exchange are rendered permissible—whatever the absolute mag-
nitude of the imposed costs and compensating benefits. Consider just two
fairly extreme examples. The first is forced participation in eyeball (or kid-
ney) redistribution pools. All members of the pool start with two healthy
eyeballs (or kidneys). If a member loses both eyeballs (or both kidneys),
some other member of the pool is required to donate an eyeball (or kidney)
for that first member. The details for membership in a given pool are set so
that, for each individual who is required to enter that pool, the expected loss
from being a member is comfortably less than the expected gain. So, the
pool is one that rational individual cost-benefit calculators would freely
enter. But imagine further that many people simply do not voluntarily sign
up for their designated pools. Liability rule attenuation seems to allow forc-
ing people into such pools on the simple grounds that the (expected) cost of
participation is more than duly compensated for by the (expected) gain.
The second example takes advantage of the fact that liability rule compen-
sation need not be direct or intended. Imagine that a number of individuals
are captured by slave raiders, transported to another continent, and held in
pretty bad slavery for the rest of their lives; but, had they not been captured,
they would have almost immediately been wiped out by an unanticipated
and horrendous disease. Even though we may blame the slave raiders for
crossing boundaries without anticipating the accompanying due compensa-
tion, under liability rule attenuation no rights of the enslaved turn out to be
violated. The enslaved would have no claims to any (further) compensation
from the raiders; and, of course, there would be no rights-violation basis for
punishing the raiders.12
11
A thorough discussion of Nozick’s view would examine the ways in which, according
to Nozick, difficulties in determining subsequent due compensation for boundary-crossings
may lead to the conclusion that many boundaries ought to be treated as though they are
protected by property rules. For a condensed discussion, see Mack (2012), pp. 100–3.
12
Railton (1985, p. 115) and Sobel (2012, pp. 46–7) also criticize liability rule
attenuation for not sanctifying the choice of the right-holder even though they themselves
would allow overriding the choices of right-holders without compensation when this is
sufficiently socially beneficial.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
is needed because (at least typically) each individual gains more in welfare
from being allowed to engage in minor intrusions than she loses in welfare by
being subject to them.13
A is worse off to the extent that he cannot do with his land what he could have done
before [minor intrusions were allowed] . . . Yet by the same token he is better off to
the extent that the same regulation binds all others (B through Z) [to allow A’s
intrusions] for his benefit. The parallel restrictions upon others become implicit
in-kind compensation for A, and likewise for all others in the group. With
compensation thus assured there is no need to undergo the expensive and pointless
process of making explicit offsetting payments, whose sole effect is to dissipate the
welfare gains generated by the change in legal rules [from those that forbid minor
intrusions]. (1979, p. 78)
Epstein’s implicit in-kind compensation position is one version of a common
response to the restrictiveness of prohibitions on minor intrusions, viz., that
surely we are all willing to be subject to a whole host of minor intrusions in
order ourselves not be required to avoid engaging in such intrusive activities.
However, it is crucial to distinguish between Epstein’s in-kind compensation
version and the elbow room version of this response. According to the
implicit in-kind compensation version, what makes minor intrusions upon
B permissible is that the actions that will be performed by B and by others if
minor intrusions are permissible will render B better off than (or at least as
well off as) B would be rendered by the actions that would be performed by
B and others if minor intrusions are forbidden. What matters is the actual
outcome for B’s utility or welfare of the introduction of the permissibility of
minor intrusions. Thus, Epstein declares that “. . . the acid test is whether the
overall scheme provides implicit in-kind compensation by allowing each
landowner an appropriate fraction of the resulting gain” (2009, p. 28, emphasis
added). In contrast, according to the moral elbow room version, what makes
minor intrusions upon B permissible is that only if such intrusions are
permissible will individuals (including B) systematically have discretionary
control over the domains that are ascribed to them as a matter of right. What
matters is that people’s rights be delineated in a way that systematically allows
individuals to dispose of their persons and property as they respectively see
fit. Thus, the permissibility of A’s engaging in minor intrusions upon B is
not hostage to B actually employing his moral liberty to engage in minor
13
Epstein could further fortify this position by pointing out that (typically) individuals
gain enormously from others being allowed to impose minor intrusions on one another.
A gains knowledge about these things called fireplaces that can keep one from freezing to
death from B is allowed to build and use a fireplace even though that deposits some ash
onto C ’s property. A is enabled to trade with B because B is allowed to drive her wagon
past C even though doing so vibrates C ’s eardrums. And so on.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
intrusions in ways that render B better off than (or at least as well off as) B
would be were minor intrusions forbidden.
Indeed, Epstein’s own statement of his position straddles the fence between
the in-kind (utility or welfare) compensation and the elbow room versions.
In the first three sentences of the passage cited immediately above, Epstein
seems to say that, although A has to allow minor intrusions on his land (e.g.,
the depositing of some smoke particles), others in turn have to allow his
minor intrusions upon them. The general restriction against prohibiting
minor intrusions leaves each better off with respect to their liberty to do as
they see fit with their property. While a recalibration of rights that construes
minor intrusions to be permissible involves a minor contraction of the
boundaries of A’s rightful domain, the associated permissibility of A’s minor
intrusions secures A’s discretionary sovereignty over that slightly retrenched
domain. Only in the final sentence in this passage does Epstein clearly express
the implicit in-kind (utility or welfare) compensation stance. Only in that
sentence is he clearly thinking about how individuals will actually act given
the recalibration of rights and is he vindicating the recalibration on the grounds
that the anticipated actions will on net advance the utility or welfare of each
of the parties.
We can see a similar fence-straddling when Epstein argues that the rights
of landowner A should not be understood as rendering it impermissible for
neighbor B, to make changes on her property that will diminish A’s view.
Beyond arguing that B’s disturbing A’s vista does not even satisfy the
“Physical Invasion Test” (1979, p. 60) for being a boundary-crossing,
Epstein offers the following consideration.
A’s claim to an unobstructed view is attractive only because it is considered in vacuo.
Yet the uniform protection of all views commits us to a set of entitlements that make
it impossible for anyone to use the land from which he might choose to look. (1979,
p. 61)
Although Epstein certainly thinks of this as an implicit in-kind compensation
argument, its focus is not on each party actually gaining (utility or welfare)
from the actions that will ensue if vista alteration is allowed. Rather, the
focal point seems to be the elbow room consideration that, only if vista
alteration is allowed, will individuals have the discretionary control over
their persons and property that we expect them to have when their
ownership rights are affirmed.
Suppose that, even if vista alteration is allowed, landowner B will never
do anything on her land that will diminish neighbor A’s existing vista; but
A will engage in activities on his land that will diminish B’s view. Under
these circumstances, implicit in-kind compensation reasoning does not
yield the permissibility of vista alteration while elbow room reason does.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
The convenience of such a rule may be indicated by calling it a rule of give and take,
live and let live . . . (Bamford 1862, pp. 32–3)
Epstein takes this passage to be an enunciation of the implicit in-kind
compensation stance. I think it is more ambiguous than that.
The majority of the justices—with whom Bramwell concurred—went out
of their way to say that A’s common and ordinary use and occupation of his
land and houses is “conveniently done” when that use or occupation is no
more than “an annoyance” to other parties (Bamford 1862, p. 30). Given this
understanding of “conveniently done,” the first sentence in the passage from
Bramwell seems to say that rights to land and houses must include rights to
dispose of those holdings even if one’s chosen action makes the residence
of another “less delectable or agreeable” (Bamford 1862, p. 30). We seem to
have a claim about the moral elbow room that must accompany one’s rights
to one’s lands and houses. If one has such rights, one must be at liberty to
use those holdings as one sees fit even if one’s exercise of that liberty makes
others’ holdings less delectable or agreeable. In contrast, the latter part of
the passage seems to compare the utility or welfare gains and losses for
individuals from their actually engaging in (non-malicious, non-wanton)
annoying conduct and being subject to such conduct. Since the losses from
being subject to the annoying conduct will be trifling and the gains from
engaging in the conduct will be more than trifling, a rule that allows each
to engage in annoying conduct is advantageous to all.
Bramwell provides a “live and let live” and a “give and take” synopsis of his
stance. The live-and-let-live synopsis fits nicely with the moral elbow room
reading. Each party has a right to live and more concretely to live by way of
discretionary control over his own person and possessions. Yet, each person’s
right of discretionary control over her person and possessions would come
to (almost) nothing if each person’s right included a right against imposed
annoyances. If each of us is to live, we each have to let everyone else live; and
this requires everyone having the moral liberty to annoy; i.e., the moral
liberty to engage in minor intrusions. This reading of the “live and let live”
synopsis fits especially very nicely with the general principle that the majority
relied in Bamford v. Turnley, viz., “No man may use his right so as to damage
another; though, on the other hand, every one has a right reasonably to use
his property, even if he should thereby annoy his neighbor”15 (Bamford
1862, p. 29). In contrast, the “give and take” synopsis fits better with
15
None of the judges who found for Bamford would allow Turnley to proceed with
his damaging fumes if he compensates Bamford. Bramwell says that a party who is
damaged has a claim to an injunction against the damaging activity. According to the
other judges who found against Turnley, if A’s operation of a tanning facility is damaging
to B, “it may be pulled down” (Bamford, p. 30).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
16
I assume that burning herself alive would be worse for B than starving—as B might
if B were too morally hog-tied to fertilize her fields or raise smelly pigs.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
argument for A and B being at liberty to engage in minor intrusions will not
go through. In contrast, the moral elbow room argument for this liberty, viz.,
self-ownership would be rendered nugatory if persons were forbidden to dis-
pose of themselves in ways that were annoying to others, will still go through.
It seems to me pretty clear that it is permissible for A to motor up his
driveway even if some noise carries over to neighbor B who never herself
in fact engages in minor intrusions upon A. Similarly, it seems to me
pretty clear that B may permissibly cause some of her ashes to land on A
or his property in the course of her self-immolation even though it is
contrary to her interest for this pollution to be permissible. Also, it is
permissible for A to motor up his driveway even if B would self-immolate
were minor intrusions permissible. If one shares these judgments, one
should prefer the elbow room account of the permissibility of minor intru-
sions over the implicit in-kind compensation argument. The elbow room
argument better accords with “. . . a picture of a free society as one embod-
ying a presumption in favor of liberty, under which people permissibly
could perform actions as long as they didn’t harm others in specified ways”
(Nozick 1974, p. 78).
Bramwell maintains that the proper delineation of mine and thine leaves
room for actions that convey low-level odors or noises or vibrations to non-
consenting others; nevertheless, the conveyance of the same odors or noises
or vibrations will count as unlawful nuisances if “wantonly or maliciously”
done (Bamford 1862, p. 32). This, I think, is probably the stance that the
friend of Lockean rights needs to take. For it allows the Lockean to hold
that “stealing a penny or a pin or anything from someone violates his rights”
(Nozick 1974, p. 75, emphasis added) without having to hold that an action
that incidentally does a penny’s worth of damage to someone’s property (or
incidentally creates a 0.1 probability of ten penny’s worth of damage) is a
rights violation. However, as Railton argues, it seems that the wanton or
malicious performance of an act of a given type will count as a boundary-
crossing only if the non-wanton and non-malicious performance of that act
counts as a (presumably less blameworthy) boundary-crossing (1985,
p. 102). The boundary-crossing character of my maliciously depositing
some ash on your backyard seems to depend upon the boundary-crossing
character of my non-malicious (and non-wanton) production of that
outcome. Thus, if the malicious minor intrusion is a boundary-crossing, so
too must be the non-malicious (and non-wanton) intrusion.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
At the same time, the lesson to be drawn from the hog-tying challenge is
that, in order for people to possess the moral liberty to do as they see fit with
their own persons and property, they must not be morally precluded from
engaging in minor intrusions that are integral to their discretionary control
over and disposition of their own persons and property. The generation of
some negative spillover effect cannot be a basis for shutting down persons’
chosen use of what is their own. If rights were delineated in such a way that
such negative spillover effects provide a moral basis for shutting down per-
sons’ chosen actions, rights would not serve their (deontic) purpose of
defining spheres of freedom within which individuals may act as they see fit.
If each person’s protected sphere were confined to conduct “which affects
only himself ” (Mill 1859, p. 11), that sphere would be vanishingly small.
The rights that we ascribe to individuals to provide them with moral protec-
tion for the exercise of their liberties must leave persons morally vulnerable
to minor intrusions if those rights are not to shrink those spheres into (at
most) pinpoints of protected freedom.
The elbow room postulate or something close to it plays a systematic role
within rights theory by guiding the reasonable delineation of rights. Here
I will cite two examples of such guidance. The first example is Locke’s argu-
ment against there being an original joint-ownership of nature. Locke
accepts Robert Filmer’s claim that, if the earth were originally the joint
property of all of mankind, the establishment of any private property—
indeed, any permissible individual use of any portion of the earth—would
require a compact among all the joint-owners that has never taken place and
will never take place (Filmer 1652, p. 234). Locke, however, takes this to be
a premise in an elbow room argument against the original joint-ownership
of the earth. For, if the earth were the joint property of all and Filmer’s claim
is correct, there would be no elbow room for individuals to exercise their
most fundamental natural right, viz., “the right everyone had to take care of,
and provide for their Subsistence” (Locke 1689, First Treatise, §87). For
(almost) any exercise of this fundamental right requires that individuals be
at liberty to acquire private property or, at the very least, to make use of
portions of the earth.17
In addition, according to Locke, private property rights themselves must
not on net diminish the opportunity of individuals “to take care of, and
provide for their Subsistence” (1689, First Treatise, §87). Moral elbow room
for this right of individuals to preserve themselves by bringing their self-
owned powers to bear on their extra-personal environment requires that, for
17
Thus, when Locke says that the earth originally “belong[s] to Mankind in common”
he simply means that “no body has originally a private Dominion, exclusive of the rest of
Mankind” (1689, Second Treatise, §26).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
each individual, others’ acquisition private property rights and the exercise
of those rights on net leave “enough, and as good” (1689, Second Treatise,
§27) for the use of that individual. It requires that the acquisition and dis-
position of property rights not on net make the extra-personal world less
susceptible to any individual’s efforts to exercise his powers in ways that serve
his ends. So, on this reading of Locke, Locke’s “enough, and as good” proviso
itself arises as a further application of elbow room reasoning. This contrasts
with the Lockean proviso that Nozick endorses which seems to allow acqui-
sitions or deployments of property rights that on net diminish the economic
opportunities of individuals as long as those individuals receive (implicit or
explicit) compensation along some utility or welfare dimension.18
The other example of elbow room reasoning concerns the right of self-
defense. Why is the use of harmful force against another agent morally
permissible when it is employed in order to block that other agent from
violating one’s rights? Why doesn’t the attacker’s right of self- ownership
morally preclude the defender from striking the attacker in order to ward
off the attack? One natural line of argument goes from the prospective vic-
tim’s self-ownership to her possession of a moral liberty to defend herself
against violations of that right and from that moral liberty to harm-inflicting
defensive acts not being boundary-crossings. If individuals are morally at
liberty to defend themselves through the use of harmful force, there must be
moral elbow room for the exercise of that liberty.19 That elbow room can
exist only if agents do not have rights against being subject to such defensive
acts. The elbow room account avoids reliance upon the idea that the attacker
forfeits in part or in whole an original blanket right against being subjected
to any harmful force. This avoidance is a good thing if one holds—as one
should—that self-defensive force may be used against blameless attackers.
For forfeiture can do no real work in explaining the permissibility of those
self-defensive acts. Invocation of forfeiture turns out to be just a back-
handed way of saying that a plausible delineation of rights does not ascribe
to (blameworthy or blameless) attackers a right against being subjected to
defensive force.20
In these cases, as in the elbow room argument for minor intrusions not
counting as boundary- crossings, I have moved from individuals possess-
ing moral liberties to act in certain ways to there being no boundaries
that stand as moral barriers to these actions. However, it might be objected
that such a bold move to the no-boundaries conclusion ignores the possibility
18
On Lockean and/or Nozickian provisos, see Mack (1995).
19
And an individual who is morally at liberty to forcefully defend herself may deputize
another to exercise that liberty on her behalf.
20
Jonathan Quong helpfully pressed me on this issue.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
of a more modest conclusion, viz., that there are boundaries that are only
protected by liability rules. I would like there to be a nice, simple, and yet
totally persuasive reason to believe that ceteris paribus an elbow room argu-
ment anchored in A’s moral liberty vis-à-vis B to do x yields the bold con-
clusion that action x crosses no boundary of B rather than the modest
conclusion that B has a right against A doing x albeit a right that is merely
protected by a liability rule. Perhaps that reason is this: to think that A’s
moral liberty is a matter of B having a claim protected by a liability rule
against A’s exercise of that liberty is to make A’s moral liberty too condi-
tional. For, it is to think that A’s moral liberty to x consists in her having that
liberty if and only if she pays a certain price for x-ing.
It is clear in the self-defense case that one gets to the bolder no-boundary
conclusion. The defender does not infringe upon any right of the attacker
that requires compensation on the part of the defender. The defender does
not have to purchase the moral liberty of self-defense from the aggressor.
The same seems true in the case of the moral liberty to use or appropriate
portions of nature. Admittedly, Locke’s argument against the original
joint-ownership of the earth presumes that this ownership would be pro-
tected by property rules—hence, any of the joint-owners may forbid any
use or appropriation of any portion of the earth. Locke does not see the
possibility of and does not refute joint-ownership protected by liability rules.
Nevertheless, as I have suggested, if the moral liberty of preserving oneself
is not itself something one has to pay for, then it seems that elbow room
reasoning that begins with that moral liberty does cut against joint-owner-
ship of nature whether that ownership be construed as protected by prop-
erty or by liability rules.
Still, it is reasonable to ask why moral elbow room reasoning in the minor
intrusion cases bypasses the modest conclusion that those intrusions do cross
boundaries and yet are acceptable if they are accompanied by due compensa-
tion. Why in these cases is it reasonable to go directly to the conclusion that
there is no boundary there to be crossed? I’ve argued that the libertarian the-
orist must reject the general liability rule attenuation of rights and, hence,
must reject the general attenuation route to the permissibility of minor intru-
sions with compensation. However, what is now on the table is the possibility
of basing the conclusion that minor intrusions are permissible on the need for
minor intrusions to be permissible if hog-tying is to be avoided.
From the need for elbow room for minor intrusions, why not draw the
conclusion that, although such intrusions do cross boundaries, these crossings
are permissible with compensation because these boundaries are merely pro-
tected by liability rules?
I think that even in the case of an agent’s moral liberty to do as he sees fit
with his person and property in ways that (non-maliciously and non-wantonly)
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
involve minor intrusions upon others the consideration that the agent need
not purchase the liberty at issue from those others has force. Consider again the
first case involving A and the over-conscientious B. A’s doing as he sees fit with
his person and property sometimes mildly intrudes upon B. Some smoke from
A’s barbeque lands on B’s acreage and some noise from his motoring up his
driveway vibrates B’s eardrums. However, in contrast to almost anyone else we
might envision, by her choice B never disposes of her person or property in
ways that are even mildly intrusive upon A. Since B does not even mildly
intrude upon A, B does not receive any implicit in-kind compensation in
the form of gains from such intrusions. So, if A’s minor intrusions require
compensation to B, it must be that B can morally require that A desist from use
of his smoker and his driveway unless A makes sufficient explicit payments to
B. That is, it must be that A has to purchase from B his liberty to use his smoker
and driveway.
However, A seems to have a strong elbow room response to B’s demand
that A purchase those liberties from B. The response is that, absent special
circumstances, to have a right over one’s person and property includes hav-
ing the moral liberty to dispose of one’s person and property as one chooses
and that appreciation for the hog-tying problem calls for a recognition that
this moral liberty includes one’s discretionary disposition of one’s person
and property even if that disposition involves minor intrusions on others. A
points out that, in light of what rights are supposed to do—he might say, in
light of the telos of rights—a sensible delineation of rights will not count
minor intrusions as boundary-crossings.
A can further point out that this delineation of rights does not unilater-
ally favor him; B also is credited with the moral liberty to dispose of her
person and property as she sees fit even if her chosen dispositions involve
minor intrusions upon A. Both parties’ rights are construed in a way that
best articulates persons’ underlying moral claim against being precluded
from living their own lives in their own chosen ways. There is, therefore, a
sense—nicely expressed in Bramwell’s live-and-let-live rule—in which B is
implicitly in-kind compensated for A being credited with the moral liberty
to engage in minor intrusions. For B too is credited with that liberty. Note,
however, that under elbow room “compensation,” each individual’s com-
pensation takes the form of the moral liberty to do as he or she sees fit with
his or her person and property without paying compensation even if the exer-
cise of that liberty involves minor intrusions upon others.21
21
Similarly, one could say that each agent is “compensated” for being subject to self-
defensive force by being herself at liberty to engage in such force. However, this is not to
be confused with self-defensive force being permissible only if the attacker is provided
with compensating utility or welfare.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
Still, I have left open the possibility that, under special circumstances,
libertarian rights may take the less stringent form of claims that are protected
merely by liability rules. Two quite different sorts of case readily come to
mind. First, there are cases of “soft paternalism.” Suppose B is subjected
without her consent to an interference that would normally count as a rights
violation. It is commonly argued that such interference may be permissible
if B would agree to it in light of the disvalue for B of the outcome that the
interference would prevent. For example, B is unknowingly about to step in
front of a speeding bus. Because of the disvalue for B of being squashed like
a bug, B would waive her right not to be yanked out of the bus’s path.
However, time constraints make it impossible for A to elicit that waiver
from B. Due to those constraints, B cannot exercise choice over whether she
will be yanked back or not. Under such circumstances, it seems pointless to
say that B’s right against such intervention includes a right to determine by
her choice whether A’s interference will be permissible or not. So it seems
that, under such circumstances, B’s right against such interference amounts
to a right not to be subject to it unless she would agree to it were such agree-
ment possible. If this is correct, then under these special circumstances, B’s
right amounts to a right to her being by her own lights duly compensated
for being subject to that interference—and being saved from being squashed
like a bug duly compensates B.22
In a second sort of case, what A has a right to and what B has a right to
are so closely connected that neither party can engage in a core exercise of
his or her right without precluding the other from engaging in a core exercise
of his or her right. For example, A cannot extract his natural gas (which lies
below his plot of land) without precluding B from retaining her natural gas
(which lies below her neighboring land) and vice versa because the gas is
part of a single pool. In situations of this sort the closest we can get to preserv-
ing the right of each to extract his or her own gas and not be subject to the
unchosen extraction of his or her gas is to ascribe to each a moral liberty to
extract gas from the pool subject to a requirement that the extractor compen-
sate the other party for the unavoidable extraction of the other party’s nat-
ural gas; A may extract B’s gas but must pay B the value of B’s extracted gas
minus the costs to A of the extraction (and sale) of B’s gas.23
22
See Steven Wall’s “Self-Ownership and Paternalism,” The Journal of Political
Philosophy, vol. 17, no. 4 (2009), 399–417. Wall argues that advocates of libertarian self-
ownership have no satisfactory explanation for the permissibility of yanking B out of the
path of the bus.
23
The detailed law governing these matters differs from state to state. In Colorado, B
has the option of sharing from the start in the financial risks of the extraction project or
not sharing in that risk and later paying a higher percentage of the costs if the project is
successful.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
24
Judicial decisions are more apt to track Lockean thinking than legislative decisions
because judges are supposed to focus on the rights of the parties who come before them
while legislators are apt to advance policies the rationale for which has little or nothing to
do with rights.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
25
Other issues would include: (i) whether an apparent trespasser does not intrude at
all because he possesses an easement; (ii) whether the complainant cannot enjoin an
already established activity that damages her when she comes to the neighborhood; and
(iii) whether the dire straits of the defendant provides him with a dispensation to engage
in conduct that would otherwise be boundary-crossing.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
Bibliography
Bamford v. Turnley (1862). 122 Eng. Rep. pp. 27–34.
Calabresi, G. and A. Douglas Melamed, A. D. (1972). “Property Rules, Liability,
and Inalienability,” Harvard Law Review 85(6), pp.1089–128.
Cohen, G. A. (1995). Self-Ownership, Freedom, and Equality. Cambridge: Cambridge
University Press.
Epstein, R. (1979). “Nuisance Law: Corrective Justice and its Utilitarian Constraints,”
Journal of Legal Studies 8(1), pp. 49–102.
Epstein, R. (2009). “Property Rights, State of Nature Theory, and Environmental
Protection,” New York University Journal of Law and Liberty 4(1), pp. 1–35.
Filmer, R. (1652). Observations Concerning the Original of Government. In: Patriarcha and
Other Writings, J. P. Sommerville (ed.). Cambridge: Cambridge University Press, 1991.
Gaus, G. (2012). The Order of Public Reason. Cambridge: Cambridge University
Press.
Locke, J. (1689). Two Treatises of Government, Peter Laslett (ed.). Cambridge:
Cambridge University Press, 1960.
Lomasky, L. (1987). Persons, Rights, and the Moral Community. Oxford: Oxford
University Press.
Mack, E. (1995). “The Self-Ownership Proviso: A New and Improved Lockean
Proviso,” Social Philosophy and Policy 12(1), pp. 186–218.
Mack, E. (2010). “The Natural Right of Property,” Social Philosophy and Policy 27(1),
pp. 53–78.
Mack, E. (2012). “Nozickian Arguments for the More-Than-Minimal State.” In:
Bader, R. and Meadowcroft, J. (eds), The Cambridge Companion to Nozick’s
Anarchy, State and Utopia. Cambridge: Cambridge University Press, pp. 89–115.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
9
Rules and Rights
Jonathan Quong & Rebecca Stone
This chapter addresses the following two questions that arise when some
agents are imperfectly motivated to comply with the demands of justice or
hold false beliefs that lead them to act unjustly:
• What rules should we devise in response to this non-compliance?
• May such rules license departures from the rights individuals would
possess under conditions of full compliance, and if so, do the resulting
rules actually alter agents’ rights?*
Problems of non-compliance don’t seem particularly troubling for theo-
ries, like utilitarianism, that focus on maximizing the aggregate or average
amount of some agent-neutral conception of goodness. This is because the
problem doesn’t fundamentally change for such theories when we move from
ideal to non-ideal conditions. Our duty is always to maximize the good as the
theory defines it. The particular prescriptions change and will be second best
under non-ideal conditions, but the underlying objective is the same. When
the theory prescribes governance by rules, moreover, agents have only
instrumental reasons to obey those rules; they have a duty to obey them
only insofar as doing so will promote the good.
For deontological theories, by contrast, the nature of the problem seems
to change dramatically in the face of non-compliance. Such theories assume
that we have moral claims against one another that are not reducible to
judgments about how much agent-neutral goodness actions or rules will
produce. Our duties to obey rules that define the scope of our moral claims
against one another are therefore not instrumentally grounded in some
* For many helpful comments and suggestions, thanks to Richard Child, Andrew
Lister, Jeff McMahan, Tom Sinclair, Hillel Steiner, Zofia Stemplowska, Chad Van
Schoelandt, and audiences at the University of Leeds, the University of Manchester, the
University of Wales, Newport, and the Oxford Studies in Political Philosophy Workshop
in Tucson, Arizona.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
Of course, critics of so-called ideal theory have other complaints too, but in this
2
3
Following (McMahan 2009, 9–10), we distinguish between rights violations and
rights infringements. The former are unjustifiable transgressions of moral rights, whereas
the latter are justified transgressions of moral rights. We use the term transgression to
cover both types.
4
There is a substantial literature on how consequentialists ought to respond to the
so-called demandingness objection, which the problem of non-compliance allegedly
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
Before proceeding any further, we should clarify a few key terms and
distinctions. First, we adopt Rawls’s account of the distinction between tel-
eological and deontological theories (Rawls 1999, 21–2, 26). Teleological
theories begin with an independent account of goodness, and then define the
right in terms of maximizing or promoting the good. Deontological theo-
ries reject this picture, either because they do not define rightness in terms
of maximizing or promoting goodness, or else because they do not define
goodness in way that is independent of rightness.5
Second, we use the terms “rights” and “claims of justice” interchangeably to
refer to moral entitlements held by individual persons that are, at least in
principle, enforceable. Our focus is thus on that domain of deontological
morality for which the term justice is usually reserved. We do not consider
whether and what moral duties and claims exist beyond this domain. To act
contrary to an individual’s legitimate rights is thus to act in a way that is,
absent some further justification, unjust by deontological standards. Likewise,
we use the term “rights-based theory” to refer to a theory that aims to provide
a deontological account of individual rights.
Third, for expositional convenience we posit the existence of a rule-
maker who designs rules to respond to the problem of non-compliance.
The rulemaker can be thought of as any decision-making body or process
that makes rules on the community’s behalf; for example, a legislature,
court, or decentralized norm-creating process. We are interested in what
the substantive content of the rules ought to be. We remain agnostic about
questions of institutional provenance and the distribution of political
power in a just society.
The chapter is structured as follows. In section 1 we outline the remedial
and rights consequentialist approaches and argue that they are flawed. In
section 2 we explain the nature of the deontological rulemaker’s decision
problem, and we argue that the costs of non-compliance can often be
treated as a burden over which individual claims can be allocated in a man-
ner consistent with widely accepted deontological principles. Section 3 con-
siders and rejects three objections to our proposal. Sections 2 and 3 thus
constitute our defense of Thesis (A): rules designed to address non-compliance
poses for consequentialist theories, but this is not our focus (e.g., Cullity 2004; Mulgan
2001; Murphy 2000). There also is a growing literature on the distinction between ideal
and non-ideal theory addressing how, if at all, this distinction ought to be drawn and
whether political theorists have been focusing a disproportionate degree of attention on
ideal theory (e.g., Farrelly 2007; Simmons 2010; Stemplowska 2008). But this debate is
also distinct from the question we focus on here.
5
It is important not to conflate the distinction between consequentialism and
non-consequentialism with the distinction between teleological as opposed to deontolog-
ical theories. In this chapter, we remain agnostic as to whether a deontological approach
like ours might be “consequentialized.”
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
can sometimes ground individual rights. These rights have the same essen-
tial features as ideal rights, and can thus properly be described as demands
of justice. In sections 4 and 5 we move on to explain and defend Thesis (B):
some rules designed to combat non-compliance may be inconsistent with
deontological principles, and as such they do not ground individual rights
but are nonetheless sometimes justified all things considered.
In this section, we present and reject the remedial and rights consequential-
ist approaches to non-compliance and explain why they are flawed.
6
Notice that the possibilities set out in (R3) don’t necessarily constitute an exhaustive
list, and so the remedial approach may allow for rules designed with other ends in mind,
so long as they don’t fall afoul of (R2). For example, the remedial approach might permit
rules designed to promote certain forms of human excellence.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
7
A defender of the remedial approach might argue that the victims of non-compli-
ance have an (ideal) right to be given the nearest attainable approximation of resources to
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
which they are ideally entitled, and that this right explains why there are duties to assist
them. But the relevant question would then be this: who bears the duties that correlate
with this right? As we indicate above, it seems clear the remedial approach must say it is
the non-compliers who bear any such duties. There is no basis, within the terms of the
remedial approach, for assuming that complying agents can be the subject of such duties.
8
Note that this is compatible with the existence of other kinds of enforceable ideal
positive duties—for example, duties to aid victims of misfortune that can’t be attributed
to anyone’s non-compliance.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
agent’s freedom can be taken away at any moment by anyone willing to pay
him adequate compensation. Someone might object that this is not so
because an agent has not been adequately compensated unless the amount
he receives in compensation leaves him rationally indifferent between the
option to freely choose whether or not to φ and being forced to φ and
receiving that amount in compensation. But the mere fact the agent is ren-
dered rationally indifferent between these two scenarios by adequately com-
pensating him doesn’t alter the fact that he no longer has a right to φ, but
only a disjunctive right [to φ or be forced not to φ and receive adequate
compensation]. The latter kind of right guarantees him a certain level of
well-being, but it doesn’t confer upon him a genuine freedom of choice
(Coleman and Kraus 1986, 1339–40). He is not free, for example, to insist
on selling his right to φ only for more than the amount of compensation
that would leave him rationally indifferent. Such a refusal might be irra-
tional. But we ordinarily think that rights confer upon agents the freedom
to take actions that are irrational or even wrong (Waldron 1981).9
If, therefore, even adequately compensated conscription is inconsistent
with respecting a plausible conception of agents’ ideal rights, it will be dif-
ficult to devise rules that can effectively combat non-compliance. But with-
out rules that require agents to assist victims of non-compliance, enforcement
will depend on the willingness of agents to volunteer their services to the
project of enforcement, and to the extent that they aren’t sufficiently forth-
coming, the consequences of non-compliance will fall haphazardly on
whomever the non-compliers decide to target—a result that stands in ten-
sion with the deontological commitment to the equality of persons.
9
For a detailed critique of the attempt to combine the libertarian commitment to
self-ownership with such a “cross and compensate” model of moral rights, see Sobel (2012).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
(C1) is identical to (R1). (C2) is the negation of (R2), and thus captures
the central difference between rights consequentialism and the remedial
approach. (C3) lists the goal of rules that justifiably infringe agents’ ideal
rights. In short, although rights consequentialism views agents’ ideal
rights as the ultimate sources of moral value, it allows, indeed requires,
that the rules be designed to allow the rights of some to be traded off
against the rights of others in order to minimize the weighted sum of
rights transgressions.10
Such an approach has a lot of prima facie plausibility when compared
with the remedial approach. Unlike his remedial counterpart, the rights-
consequentialist rulemaker can forcefully intervene on behalf of the
innocent victims of non-compliance by forcing others to assist in his
enforcement efforts so long as the result is a lower number of rights
transgressions overall.
Notice also that rights consequentialism can be easily modified to accom-
modate a concern with equality. (C3) states that the goal is to minimize
the weighted sum of rights transgressions, but plausible versions of rights
consequentialism can be constructed in which this purely aggregative cri-
terion is replaced or qualified by distributional criteria that require that
rules be designed to ensure that rights transgressions are not too unevenly
distributed.
The problem with rights consequentialism is that it isn’t truly compatible
with the ideals that underpin most deontological theories. Since (C2) per-
mits rights transgressions for the sake of the greater good, rights consequen-
tialist rules don’t generate claim rights of any kind.
Thus, even though (C1) implies that the greater good is measured in
terms of agents’ ideal rights, the resulting rules don’t really respect the rights
on which they are grounded. Agents’ duties to obey the rules are impersonal
or agent-neutral duties that aren’t grounded in particular agents’ claims.
Given widespread non-compliance, the rules will regularly require rights
infringements, with the result that agents will typically have many imper-
sonal duties to infringe the claim rights of others. Thus, it is not clear what
is really left of agents’ claim rights. They continue to make genuine claims
on agents in the interstices of the rules. But when the problems of non-com-
pliance are severe, the freedom that is implied by the interstitial respect of
agents’ claims becomes less and less meaningful. In Nozick’s terms, rights
consequentialism leads to a “utilitarianism of rights” that requires us “to
violate someone’s rights when doing so minimizes the total (weighted)
amount of the violation of rights in society” (Nozick 1974, 28–9). Thus, it
10
Amartya Sen (1982) famously proposed that minimizing rights transgressions could
be one of the goals of a consequentialist theory.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
11
And, as we explain in sections 4 and 5, even if there is sometimes a weighty conse-
quentialist justification for infringing basic individual rights, our view differs from rights
consequentialism in identifying this as a lesser evil justification, and not one that is
grounded in an appropriate conception of justice.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
claims of justice on the members of the community. That is, they set limits
on what the state may do to its members in order to prevent and remedy
rights violations, while also ensuring that members have a duty to other
complying members of the community to contribute to the state’s enforce-
ment efforts up to those limits. This is because they provide agents with a
way of fulfilling their duties to assist victims of non-compliance in an effi-
cient and fair way while preserving a protected sphere for each agent in
which he is free to do as he pleases.
Importantly, we can think of agents’ resulting loss of freedom as some-
thing that has been stolen from them by non-compliers. The non-compliers
ought to return citizens’ stolen rights to them by ceasing their non-compliance
and compensating them for their past acts of non-compliance. But when
the non-compliance is recalcitrant, the community has to prevent the costs
of non-compliance being born arbitrarily by the victims of the non-compliance,
and when it does so in a just way—that is, in a way that respects the free-
dom and equality of persons—new moral rights and duties between the
complying members of the community and between the complying mem-
bers and the state arise.
Is there any room for additional rights-generating rules that might more
effectively tackle the problem of non-compliance? Unlike rights consequen-
tialism, our approach won’t permit the state to take more from certain citizens
simply in order to more efficiently deal with the problem of non-compliance.
Suppose, for example, that non-compliance arises predominantly from a
particular ethnic group in the community, thus making racial profiling an
efficient state response from a rights-consequentialist standpoint. Rights
consequentialism would naturally permit such a policy since it allows non-
compliance to be tackled more effectively at lower cost; that is, members of
the community who were not part of the targeted group could cede fewer
of their ideal rights to the community, and on average the rules would tend to
better preserve agents’ ideal claims.12 Our approach, by contrast, won’t permit
such rules to alter individual moral rights if the only justification for the rule
is its efficiency. Such rules essentially end up using one set of citizens for the
benefit of others, and so fail to respect freedom and equality.
12
More complex versions of rights consequentialism can, of course, take distribu-
tional aims directly into account, and thus might not be as vulnerable to the objection
raised in the main text. But these versions of rights consequentialism still differ from our
own approach in at least two important respects. First, for rights consequentialists, the
resulting rules do not ground individual moral rights—they are simply instruments to
achieve the best possible distribution of ideal rights transgressions. Second, the rights
consequentialist is committed to the proposition that all rights are fungible, whereas our
account is not committed to this idea. We remain open to the possibility (indeed it seems
very likely) that once the principles of liberty and equality are properly specified, it will
turn out that there is no general fungibility of rights.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
13
G. A. Cohen makes a related distinction between applied rules of regulation that
serve justice in particular and applied rules that are impure in that they serve other prin-
ciples alongside justice (2008, chapter 7).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
of freedom and equality, since any plausible rule governing such ‘lifeboat’
scenarios will allow some agents to die so that others may live.
But this response does not vindicate the original objection. We don’t
deny that some changes to empirical assumptions may make it impossible
to design rules that effectively instantiate deontological principles. We make
the weaker claim that the costs of non-compliance do not always have this
effect, for sometimes the costs of non-compliance can be distributed in
accordance with deontological principles. Moreover, it will be very difficult
for the proponent of the objection to resist this conclusion once she grants
that not all changes to the empirical assumptions of ideal theory necessi-
tate departures from the deontological perspective on justice. If our critic
grants, for example, that moderate changes to assumptions about scarcity
or human psychology need not threaten the pedigree of the resulting theory,
why should moderate departures from the assumption of full compliance
be any different?
(1) Albert rescues Betty, resulting in Eric’s death; (2) Albert rescues Eric and
Carl does nothing, resulting in Betty’s death; and (3) Albert rescues Eric and
Carl rescues Betty, resulting in no deaths. ROII tells us that because option
(1) is possible, Carl cannot be required at the bar of justice to rescue Betty.
But this seems absurd. Option (1) is much worse than option (3), and Carl
knows that Albert won’t choose it. It’s hard to understand how the existence
of such an option—one that won’t eventuate and is much worse overall—
can explain why justice permits Carl to allow Betty to drown.
Finally, ROII is likely inconsistent with practices and institutions that are
widely accepted as essential in a liberal democratic society. Any such society
will have civil and criminal enforcement systems. These institutions are
invariably funded out of general taxation: it is not simply non-compliers
who fund the police force and the courts; everyone bears the costs of sus-
taining such institutions. It is implausible to hold, as the proponents of the
responsibility objection presumably must, that compliers cannot be required
to support these institutions as a matter of justice whenever it is theoreti-
cally possible to impose the costs of sustaining those institutions on the
non-compliers.14
The proponent of the responsibility objection might protest that the
criminal justice system is in large part designed to deter non-compliance—
that is, to prevent non-compliance before there exist any non-compliers
who could be required to bear the relevant costs. And once the costs of
supporting the criminal justice system are construed in this way, it is con-
sistent with ROII to require everyone to share in the costs of deterrence,
since these are not costs for which anyone is morally responsible.15
This response, however, is problematic. First, although there is much con-
troversy about the underlying justification of the criminal justice system, not
all of the costs associated with the system can be justified in terms of deter-
rence. For example, the reason indigent criminal defendants ought to be
provided with free legal counsel is not explained by an appeal to deterrence,
yet we take this to be a cost created by the fact of non-compliance to which
all citizens can justly be expected to contribute.16 Second, setting the first
14
It is true that non-compliers could often be required to pay court costs, but the
question here is who may justly be required to bear these costs when forcing the non-com-
pliers to bear them is not feasible.
15
Even in an ideal world, some enforcement mechanisms may be needed to solve
assurance problems, in which case a citizen could ideally be required to bear his share of
the associated costs, but we think it is clear that existing enforcement regimes do more
than merely provide assurance.
16
To the extent that a fairer process is also more accurate and so ensures fewer mis-
takes, deterrence objectives may be served by giving indigent defendants a right to coun-
sel. But presumably considerations of fairness rather than efficient deterrence are the
primary reason why justice requires some form of indigent representation.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
point aside, it is implausible to draw a bright line between the ex ante costs
of deterring non-compliance and the ex post costs created by non-compli-
ance. The problem is that many of the policies and actions involved in deter-
ring non-compliance occur only once some people have made it clear that
they intend not to comply. Sending police officers to the home of a woman
who credibly reports that she fears her husband may assault her, for example,
is a cost that arises only because the husband is responsible for engaging in
threatening behavior. More generally, the costs involved in protecting peo-
ple’s rights before a right has been violated but after the potential violator has
given the potential victim grounds to be fearful is a significant feature of
contemporary law enforcement. The distinction between costs for which
no one is responsible and costs that some people are morally responsible for
creating thus does not track the distinction between deterrence costs and
post-non-compliance costs.
enforceable duties, but unlike ours, her argument is limited to cases where some people
are under a duty to rescue others in urgent or dire need. She makes no claims about how
the costs of non-compliance ought to be distributed in other situations.
20
Note that our claim is not that rules that meet the three conditions listed above
necessarily conflict with the deontological principles: many such rules can be consistent
with the principles of equality and freedom. Rather, our point is that meeting the condi-
tions set out above is no guarantee that the rule will be consistent with those principles.
Also note that Thesis (B) does not depend on endorsing this particular account. So long
as one accepts that under some conditions impersonal considerations can outweigh or
override individual claims of justice, then one can endorse Thesis (B) even if one rejects
our particular test (e.g., by rejecting the anonymous Pareto test and replacing it with a
more aggregative principle).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
be more easily enforced; and (ii) would result in a state of affairs that would
be much better for all compliers than the status quo or any attempt to force
non-compliers into compliance. Then, even though compliers will be vul-
nerable to having their rights infringed by conduct that is licensed by the
rule, they might not be able to declare the rule to be unjustified all things
considered, since they all do much better under the rule than under any
feasible alternative.
But while they may be all things considered justified, such rules do not,
on our view, generate moral rights because they represent a capitulation to
the de facto power of non-compliers to flout the conduct that is ideally
required. Such rules are not designed to ensure that compliers share fairly in
the costs of combating non-compliance taking the behavior of the non-
compliers as a given. Rather, they are designed to get non-compliers to
behave in a way that is less unjust than their current behavior. This differ-
ence of moral aim explains why these kinds of rules do not generate legiti-
mate moral entitlements.
It may help to illustrate this point with a simple example. Suppose there
is a religious society where a substantial percentage of the current popula-
tion believes that female adultery (but not male adultery) is punishable by
death, and all attempts to persuade these people that female adultery is not
a punishable offense, or to coercively prevent them from acting on this
erroneous belief, fail. Thus, the status quo is that many women are mur-
dered on grounds of adultery. But suppose the rulemaker could get the
population to endorse and conform to the following rule: all female adul-
tery is a serious wrong, and punishable by two years in prison, but no greater
punishment is permissible. Let’s suppose that any attempt to legally
entrench a lesser punishment (or to equalize the treatment of men and
women) would be met with fierce resistance and would lead to a return to
the status quo, but the population’s religious beliefs are just flexible enough
to accommodate the idea that two years’ imprisonment is a sufficient pun-
ishment for female adultery. Under these conditions, the rulemaker might
be justified in proposing and implementing this rule, yet it seems clear that
this rule is not of the sort that actually modifies individual moral rights. The
rule might have a sufficient consequentialist justification, but because that
justification involves capitulating to the beliefs of non-compliers rather
than fairly sharing the costs of combating non-compliance, it cannot alter
individual moral rights.21 It is a lesser evil, but choosing the lesser evil is not
21
The legal rules governing conduct in warfare might have the same structure as the
rule in this example. Jeff McMahan (2004, 730–3; 2008; 2009) has suggested that these
rules might have a valid consequentialist justification but don’t reflect persons’ moral
rights, since justice ultimately distinguishes among combatants according to whether
they have justice on their side, whereas the rules of war must capitulate to the reality that
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
both sets of combatants will typically believe that they are fighting for a just cause, such
that rules of war treating combatants asymmetrically will be doomed to failure.
22
Cf. Maryland v. Shatzer, 559 U.S. 98, 103 (2010) (describing the measures adopted
by the US Supreme Court in Miranda v. Arizona, 384 U.S. 436 (1966), as “prophylactic
measures to protect a suspect’s Fifth Amendment right from the ‘inherently compelling
pressures’ of custodial interrogation” (emphasis added)).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
23
In the following section we confront the objection that if an institutional rule is
morally justified then the rule must also determine the content of individual moral
rights.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
For a defense of the view that rights can conflict, see Quong (2012).
24
objection to Jeff McMahan’s distinction between the morality of war and the laws of war.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
rule would in fact tend to encourage violence. But these sorts of considera-
tions are not the kinds of considerations that can alter moral rights within
the confines of a deontological theory. The rule in question does not instan-
tiate principles of freedom and equality—the costs of non-compliance
might not be equitably distributed and freedom is not adequately pro-
tected—and so even if we believe that the rule has a sufficient justification,
it does not follow that anyone’s moral rights have been altered. This is thus
a case where persons’ moral and legal liberties come into conflict.
In cases like this it makes more sense to say that there is a sufficient jus-
tification for a rule that will result in a number of rights infringements. As
we said before, this is a lesser-evil justification; choosing the least bad
option is not the same as acting in accordance with the demands of justice.
This particular justification grants state officials the permission to prevent
black persons from entering certain parks, but black persons continue to
have the moral liberty to enter these parks: their moral liberty rights remain
unaffected.
In sum, some valid rules may have consequentialist or lesser-evil justifica-
tions, but unless those rules also instantiate the relevant deontological prin-
ciples, they don’t alter anyone’s moral entitlements. Such rules should be
understood as legally licensing certain infringements of moral rights for the
sake of guaranteeing a greater level of protection overall. These rules allow
the non-compliers to gain at the complying agents’ expense, and giving
such a windfall to the non-compliers is a pragmatic response to the problem
of widespread and recalcitrant non-compliance: it is not a just distribution
of the costs of non-compliance among those who comply.
In claiming that rules can sometimes justifiably override or ignore indi-
vidual moral rights, we don’t purport to be saying something that hasn’t
been said by many others. Our point is rather to emphasize the importance
of avoiding two tempting, but seriously mistaken positions. One is the view
that rules that respond to the problem of non-compliance can never alter
the fundamental moral entitlements granted to persons at the level of ideal
theory. The other is the view that justified rules always affect the content of
individual moral rights. Neither of these views is correct. Whether or not a
rule alters moral rights depends on the type of moral justification for the
rule. When the justification is grounded in deontological principles, the
resulting rules do determine moral rights, but rules can be justified in other
ways, in which case they don’t.
A rule is morally justified when all of its associated features (including the
enforcement of the rule) are morally justified. So once we know that a rule
is morally justified, so the objection runs, there is no point in asking whether
the rule generates moral rights. The answer is practically inert—devoid of
action-guiding implications.
We can see this objection is mistaken, however, once we reflect on the
question how ought we to respond to those who break the rules. Let’s com-
pare the example of the racially segregated parks with the case (described at
the end of section 3.3) where a small tax is imposed to fund better street
lighting to deter would-be rapists. How should a fellow citizen respond to
rule-breakers in each case? Should we respond in the same way to a black
person that we see entering the “whites only” park as we do to a person who
avoids the taxation requirement in the street lighting case? Clearly we ought
to respond to these two rule-breakers very differently. In the segregation
case, it would not be appropriate for bystanders to condemn or sanction the
black person for entering the park (even if state officials may prevent him
from doing so) at least so long as it is apparent that his doing so will not
trigger any violence. But the same is not true of the person who avoids
paying the tax for the additional street lighting. In the latter case, reactive
attitudes of blame and perhaps even sanctions are appropriate responses to
the rule-breaker on the part of all citizens. Thus, determining whether a rule
generates genuine moral rights is not practically inert. When a moral right
has been violated a different response is warranted.
Moreover, compensation is often appropriate when a genuine moral right
has been violated. This is less likely to be true when a rule has been broken
but no moral rights have been violated.26
Both this objection and the previous one misunderstand the relationship
between the moral justification of rules and moral rights. Whether a given
rule has a sufficient moral justification depends on a wide variety of consid-
erations including, perhaps, impersonal values and also, importantly, what
other rules are currently feasible. The justification of moral rights depends
on a narrower set of considerations. Whether a proposed moral right is in
fact a moral right depends on whether the justification for that right correctly
26
May state officials permissibly punish someone who violates the rule racially segre-
gating the parks? There might be extreme cases where the state would be permitted to
imprison such a person for consequentialist reasons (e.g., because failure to do so would
lead to the sort of violent conflicts the rule was intended to avoid). However, it would be
a mistake to equate such imprisonment with “punishment” in the standard sense of that
word, since the standard sense implies the criminal is liable to such punishment (i.e., has
forfeited his right by virtue of a wrongful act), but that is exactly what we are denying
occurs in such a case. So the state might, in extreme circumstances, be morally permitted
to do something to the rule-breaker that looks like punishment, but it would not be
punishment in the standard sense.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
6. CONCLUSION
Bibliography
Cohen, G. A. (2008). Rescuing Justice and Equality (Cambridge, MA: Harvard
University Press).
Coleman, J. and J. Kraus (1986). “Rethinking the Theory of Legal Rights,” Yale Law
Journal 95, 1335–71.
Cullity, G. (2004). The Moral Demands of Affluence (Oxford: Oxford University Press).
Farrelly, C. (2007). “Justice in Ideal Theory: A Refutation,” Political Studies 55,
844–64.
McMahan, J. (2004). “The Ethics of Killing in War,” Ethics 114, 693–733.
McMahan, J. (2008). “The Morality of War and the Law of War.” In Just and Unjust
Warriors: The Moral and Legal Status of Soldiers, David Rodin and Henry Shue
(eds) (Oxford: Oxford University Press), 19–43.
McMahan, J. (2009). Killing in War (Oxford: Oxford University Press).
Miller, D. (2011). “Taking up the Slack: Responsibility and Justice in Situations of
Partial Compliance.” In Responsibility and Distributive Justice, Carl Knight and
Zofia Stemplowska (eds) (Oxford: Oxford University Press), 230–45.
Mulgan, T. (2001). The Demands of Consequentialism (Oxford: Oxford University
Press).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
10
What Is Wrongful Exploitation?
Thomas Christiano
1
See Christiano (2013).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
3
See Aquinas (2003), ST II-II Q. 77, Art 1 for the principle of equality in exchange.
4
See Marx (2003), p. 392 for an account of unequal exchange between capitalist and
worker based on the labor theory of value. More recently, see Roemer (1985). And see
Wertheimer (1996).
5
See Aquinas (2003).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
rescue case, Aquinas would say that the rescued person’s life is not given to
him by the rescuer but is saved due to the special circumstance the rescued
person is in. The rescuer did undergo some cost in helping the rescued per-
son and that is what the rescuer can be said to give the rescued person but the
life or even the great utility that the rescued person received was not given to
him by the rescuer. A higher price than the market price for the cost of the
rescue is greater than the value that the rescuer contributed.
This is an intriguing argument and there is something right about it but
the basic problem with it seems to be that the value involved in the cost is
itself a function of the “circumstances affecting the buyers.” It is a function
in part of the supply and demand for the object. So, it is hard to see how the
market price is giving us a valuation of the product that is independent of
the circumstances of the buyer.
Once we reject the objective measures of value involved in the Aristotelian
and labor theory approaches, the competitive market price seems like a good
candidate for measuring value. But if we take the competitive market price as
the measure of value, there are a host of difficulties to be worried about. One,
it is often hard to know what market we are talking about. Does it have to be
world market prices or local prices? Is price discrimination generally wrong?
What of the market of like objects? Is that relevant? Two, there often are no
competitive markets for the goods involved. The creation of international
treaties is sometimes said to be exploitative but it is hard to characterize it
in terms of markets. Three, sometimes competitive market prices can be
exploitative as well. Intuitively, at least, Marx is on solid ground when he
asserts that the competitive market price for labor can be exploitative even
when there are a number of employers, as long as there are a lot of unem-
ployed people, the workers are not organized and the alternative to employ-
ment for most workers is very bad, while the alternative to not employing a
worker for an employer need not be bad since there are many others.
The price, supposing the groups of employers and employees are suffi-
ciently well segregated (due to sharp differences in education, wealth, etc.),
will be quite low and intuitively it would seem that it could be quite wrong-
fully exploitative.6 And sometimes the very idea of offering a price for a
certain kind of good suggests exploitation, as in the case of prostitution. We
needn’t agree with every one of these cases to see that the idea of exploita-
tion as unequal exchange measured in terms of market price is not giving us
a necessary condition of exploitation.
There are three other classes of cases that might suggest a more general
criticism of the substantive views. The first is the case of irresponsible
6
See Arneson (1996). See also Roemer (1985), p. 35. Wertheimer acknowledges some
of these difficulties, which is why he is tentative about his proposal.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
exchange. Suppose one person wants to buy a car but is not willing to put
any time into determining what a good car is or what an appropriate price
is and so goes into a store willing to get the first car at the first announced
price and does so. Now suppose that the usual job of the car salesman is to
bargain, fully expecting that the buyer will haggle and the price will come
down. Instead, the buyer accepts the first announced price and the deal is
made right away. The salesman even says that they can talk more about the
price but the buyer just waves him off. Here, the buyer has bought a car at
a price that is greater than the market price or the cost of getting rid of the
car or any objective price that might be thought reasonable. But it is not at
all clear that the buyer has been wrongfully exploited even though the deal
may be unequal in some important way. Here, it seems that features of the
process are driving our judgment that the seller is not taking unfair advantage.
The buyer has every opportunity to make a good deal and does not avail
himself of them.
If irresponsible or heedless participation in exchange can legitimize an
exchange even though the exchange is unequal, it appears that substantive
unfairness in the outcome is not a sufficient condition of wrongful exploitation.
This is because a feature of the process, the irresponsible action of the buyer
and the reasonably conscientious actions of the seller, can defeat the idea that
there is unfairness in the agreement even if the price is obviously problematic
by normal market standards. Unequal exchange seems insufficient to justify a
judgment of exploitation.
It seems unnecessary as well in many circumstances. For instance, if
the salesman manipulates the buyer into buying a car at a good price and
the buyer did not want to do this, it seems like there may be exploitation
here as well even though the buyer paid a good price. Here, process seems
to be driving our judgment that there is wrongful exploitation despite
equal exchange.
Two other examples suggest that unfairness in the outcome is not neces-
sary to exploitation either. First, some kind of deception or emotional
manipulation to get someone to engage in an exchange that they do not
want can undermine an exchange even if it is for a fair price. Suppose that
B does not want to sell her car to anyone. But B loves A though A does not
love B. A gets B to sell her car to him, for the market price (or any price that
seems fair), by saying how much he needs the car and intimating that he
might love B if she sells him the car. Or A relies on the existence of some
kind of emotional quirk in B to get B to sell him the car at the fair price.
B becomes flustered under certain circumstances at which point she can be
manipulated into doing something she later regrets. A knows this and
pounces and thereby gets B to sell at the market price. Another case is one
in which A enters into some kind of convoluted exchange with B which has
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
the effect that B sells her car (without entirely understanding this). She sells
it at the fair market price. But I think we would still want to say that A has
exploited B in these cases.
To be sure, we might be able to apply Aquinas’s idea here that the seller
may sell an item at a higher price if the seller values the item at that higher
price. But even here we may run into trouble. For suppose that B is willing
to sell her car at some price but not to A. A very much wants B’s car. But
A nevertheless manages to get B to sell her car at the price she wants to
sell it through one of the machinations mentioned above. It still seems
that A has exploited B. Now, there may be a further measure of the value
of B’s car that somehow includes the value of selling it to one person rather
than another or on one day rather than another or any set of details that
determine the circumstances under which B is willing to sell her car. But
the standard of value for determining equal exchange seems to become
elusive in these circumstances. For any particular standard it would seem we
can set up an example that suggests that there can be exploitative exchange
even while there is equality. Here, it seems that we have a case in which
we can have an exploitative agreement even though the price paid for the
item is a fair price.
The substantive outcome of the transaction is fair in whatever sense one
wants to think of it but the transaction is exploitative. To be sure, these are
cases of manipulation but they are also cases of exploitation, since one per-
son is benefitting at another’s expense.
Another example of exploitation that Alan Wertheimer discusses seems
to be a counterexample to the thesis that unfairness in the distribution of
goods that arises from an exchange is necessary to an exploitation claim.
The example is one of a psychotherapist engaging in sexual relations with a
client.7 One might think that the psychotherapist is exploiting the client in
this context but it is hard to see that this is the consequence of an improper
distribution of the benefits of the relationship. No doubt there is a rule
with some moral force against sexual relations between therapist and cli-
ent, which is there to protect the integrity of the therapeutic process and
thus in some sense to protect the interests of the client and the therapist.
The therapist violates that rule by proposing sexual relations with the cli-
ent. The therapist is exploiting the client because she is relying on a weak-
ness in the client, which the therapeutic process is designed to help remedy,
to acquire some benefit for herself.
So, against the substantive accounts we can say that they fail to provide
necessary conditions of wrongful exploitation because some cases of
exploitation seem to involve exchange at fair prices, however one wishes
7
See Wertheimer (1996), Ch. 6.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
8
See Valdman (2009), pp. 3, 9. Another account that is close to this is in Arneson
(1982).
9
Valdman characterizes an unacceptable alternative as one in which urgent needs are
not met or in which it is not possible to lead a minimally decent life.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
amounts to. We have seen above that this is hard to pin down. My thought
is that we have to import another idea here to explain it, but I will discuss
this in what follows. In any case, since the view does not seem to capture
necessary conditions, it does not give us an account of what exploitation is
or even what makes exploitation wrongful more generally.
elaborating can explain the idea that the exploiter wrongs the exploitee. It is
not merely the case that something bad or problematic has happened; a
wrong has been done to the exploitee.10 A second related issue is to be able
to explain the specific character of the wrong of exploitation. We will see
that this is an issue for the view I will defend. A third desideratum is the
avoidance of what Pogge calls the moral counterproductivity of such a prin-
ciple.11 In particular, I am concerned with avoiding a principle that would
prohibit opportunities to act that make everyone better off while permitting
people to avoid acting in a way that makes everyone better off. The usual
accounts of exploitation seem to imply this because they suggest that some
mutually advantageous and consensual agreements may be wrongful
exploitation and thereby prohibited. But they also suggest that the exploiter
may withdraw from agreements altogether in preference to a non-exploita-
tive agreement. The consequence of this kind of approach may be that
many persons are made worse off if people act in accordance with their
moral duties but do not do much to act morally beyond what they are
required to do. For instance, it may turn out that many people will be left
unsaved if there is a requirement that the savers do so for free or very little
recompense and there is a permission not to save or put oneself into the
position of being able to save another.
This might seem to protect a person who is being saved from extortionate
demands, but it may also create incentives in people to avoid situations in
which they can help. While if one can demand anything one wants, more
people may put themselves in the position of helping others and thereby
more people may be helped. It is an aim of my theorizing to avoid the two
possibilities of saying that exploitative action is pro tanto permissible on the
one hand or on the other hand saying the exploitative action is impermissi-
ble but one may avoid engaging with the other altogether, thus ensuring the
other is worse off.12
My concern here is grounded in my general sense that moral principles
are generally guided by a concern to advance the interests of human beings.
But I think one may be concerned to avoid the counterproductivity in this
case by a less controversial concern, which is that a principle that is meant
10
I think this is a worry for Steiner (1984).
11
See Pogge (2009), p. 115 and Wertheimer (2010).
12
We will see that the avoidance of counterproductivity in moral principles is quite
different from acceptance of what Wertheimer calls the “non-worseness” principle. This
principle says that if B consents to, and is made better off by, transacting with A then the
transaction cannot be morally worse than A’s not transacting with B. So, if A is permitted
not to transact with B, then any mutually advantageous and consensual transaction is
permissible. See Wertheimer (1996), p. 289. Like Wertheimer, I find the non-worseness
principle quite unintuitive but it is asserted by Zwolinski (2007), p. 708.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
to protect the interests of a person not be such that it actually has a ten-
dency to set back the interests of those persons. Avoidance of this feature is
on a par, I think, with avoidance of the implication of leveling down from
a principle of equality.13
I will be content if it can be shown that a principle that prohibits unfair
advantage taking does not, through a combination of requirements not to
exploit and permissions not to help, normally set back the interests of those
it is supposed to be protecting. There may be particular cases in which
everyone’s interests are set back by the application of the principle, but if
these are the exceptions rather than the norm, that may be sufficient to
vindicate the principle under discussion. The moral status of exploitation is
that it is pro tanto wrong, though it may under some circumstances be all
things considered permissible. Indeed, this condition is one that I think
offers support for the view that I will defend in what follows.
3. EXPLOITATION
13
I have tried to show that the best understanding of the principle of equality in dis-
tribution of goods implies that the principle of equality is opposed to leveling down. See
Christiano and Braynen (2008).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
There is no circularity here. The low cost is specified independently of the transac-
15
tion. It is an admittedly vague but very important component of the duty of rescue. See
also Snyder (2013).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
Here, I will explain this account by laying out some of the main examples
of exploitation.
16
See, for example, Mill (2008), Book V, Ch. XI for a discussion of the proper extent
and limits of laissez-faire in markets.
17
See Christiano (2013).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
4.1 Rescue
In the standard rescue cases, one person (A) is confronted with another
person (B) who needs to be rescued or face a very high risk of severe harm
or death. And by hypothesis, B has no further option because no one aside
from A can help him and because he needs help immediately. What I want
to say is that in the normal case this implies that in the context, A may not
morally withdraw from the interaction unless there is a great cost to him. A
has a duty to rescue B at fairly low cost to himself. This makes it impermis-
sible morally for A to bargain to full advantage with B over the conditions
under which the rescue is to take place. To be sure, A should receive some
fair recompense for helping B, if the costs are substantial. But A must figure
out how to determine fair recompense without hard bargaining with B.
If A does bargain hard with B under these circumstances, A is in effect
saying to B that he will not help B unless B pays a sum that is greater than
necessary to assure A that his cost is low. If A succeeds in this, A is exploiting
B. But he does this by violating the duty to B.
The above case is a case of the ordinary duty to help others at low cost to
oneself. But there are other cases of helping others that do not require such
a low cost. In the case of persons who help others as professionals, such as
doctors or lawyers or tow truck drivers, these may in fact charge a higher
price for their services since their livelihoods and expertise depend on such
payments. There are duties limiting what a professional can ask in these
circumstances but the duty is not limited to making sure that the profes-
sional suffers a low cost. Indeed, it is important that the professional be a
beneficiary in these kinds of cases so that there are adequate incentives for
people to engage in this kind of work. I think these kinds of cases motivate
the fair price norm of exchange and may be the motivations for the substan-
tive views. The reason why is that these cases involve a division of labor with
a great deal of asymmetry of information between client and professional
and they often involve a kind of monopoly power. A society benefits greatly
by the existence of this division of labor and wants to ensure that there are
enough people who will occupy these positions in the division of labor.
Hence, it is important for a society to provide adequate compensation for
these services. Yet, because of the asymmetries of information, it is hard for
the client to determine what the appropriate price is supposed to be. The
client is in a potentially very vulnerable position in need of these services.
Hence, there is a need to have a sense of a fair price to be paid for the ser-
vices. This price may be defined in terms of a kind of market price deter-
mined by an independent appraiser (as in the case of houses) or as
determined by a professional association. But one must note that important
as these cases are, they are not the only kind of case and so the appearance
of a general need for a conception of equal exchange is illusory.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
18
See Vrousalis (2013), p. 132.
19
See Valdman (2009), p. 4. Valdman also complains that the duty to aid account
cannot explain why a very excessive price is worse than a somewhat excessive price since
both demands would violate the duty to aid. But I don’t see why we can’t say that the duty
to aid is more seriously violated the more excessive the price.
20
Another objection that I have not had enough time to consider so far is from A. J. Julius:
“A threat to do what’s independently permissible can count as wrongfully coercive. I should
not announce that unless you do my chores I’ll sell a couch of mine that holds fond
memories for you. My broadcast of an intention to report the affair, when it figures in a
plan for manipulating the adulterer’s conduct, seems wrong even on the hypothesis that
I owe such a report to the spouse.” My inclination is to think that the moral environment
is more complicated than Julius is suggesting. We have moral reasons to take the person’s
fond memories into account and not merely to use them. See Julius (2013).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
22
I defend this principle of equality of opportunity as a principle of fair exchange: see
Christiano (2013).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
I have not yet arrived at an account of exactly how to determine how much the
24
employer might owe in any of these circumstances. My guess is that while the employers
usually owe significantly more than the market wage is in these cases, there is a significant
amount of uncertainty here and that there is no exact wage or set of working conditions
owed. This would seem to conform to the view that I am defending.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
C and D. Much background unfairness may still remain. But in doing its
fair share in alleviating this unfairness C is engaging in non-exploitative
behavior even if unfairness in the circumstances remains.25 Here, we see
how the key distinction between the idea of background or structural
unfairness and unfair advantage taking makes a difference to our discussion.
But there may still be some discomfort here. For what is required to defeat
a charge of wrongful exploitation on my analysis here is not that one pro-
vide the worker with some particular amount of good, say a living wage, but
that one do one’s part in helping the worker. What this will amount to
cannot be determined in advance and may sometimes fall short of some-
thing like a living wage.
Here too one might worry that this account will discourage people or
groups like C from employing people like D because we have argued that C
must carry a heavier cost than a simple market interaction would require.
This would be a worrisome implication.
D would plead with C to employ him even at the exploitative wage. D
would be worse off if C were to attempt to avoid the transaction (as would
C). Again, we would have something like a leveling down worry here. The
very moral principles that are designed to help the vulnerable worker would
actually damage the worker. And I think that many accounts of exploitation
either imply this leveling down or end up accepting in some grudging way
the permissibility of exploitation.
But here too the account I am offering gives us a more satisfying approach
to these problems. I want to argue that the duty in virtue of which C’s mar-
ket behavior towards D is exploitative is one that C has independent of the
market transaction. C cannot escape the cost by simply avoiding the devel-
oping world. In effect, there is a cost that C must carry no matter what C
does and so this cost should be understood to be the moral baseline in terms
of which C ought to be reasoning about the case. C cannot say, “In order to
avoid the cost of non exploitative transactions I will simply avoid transac-
tion with D.” This is because C must suffer this cost anyway as a matter of
moral requirement.
To be sure, it is permissible for C to avoid this particular transaction, but
the view I am trying to defend here suggests that C will not benefit from
25
Hence, I think Alan Wertheimer’s worry that it would be “unreasonable to expect
the better-off party to repair those background conditions [of social injustice] by adjust-
ing the terms of a particular transaction” can be assuaged once we see that the duty of
repair is not to solve the whole problem but to contribute one’s fair share to the solution.
This can amount to quite a bit when we consider all the participants in the chain of
production but it is not going to be unreasonably demanding on any particular partici-
pant. See Wertheimer (1996), p. 234. See also Young (2004) for a nuanced discussion of
the responsibilities developed-country persons have to developing-world persons.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
avoiding the transaction because C (and all the other wealthy participants
in the supply chain including the consumers) have imperfect duties to help
so that they do not avoid the costs by avoiding the transactions. This is a
way in which the view I am defending here avoids counterproductivity
since avoidance of transactions do not benefit persons. At the same time,
the view I am defending rejects the “non-worseness” principle, which asserts
that a mutually beneficial and consensual transaction cannot be morally
worse than the avoidance of transaction. In any case, C may permissibly
avoid the transaction with D, but C cannot avoid the cost.
One last note to this kind of case: C may not discharge its duty of doing
its fair share of enhancing the opportunities of human beings generally in
the world (by say contributing to a fund that does this) and then exploit the
hell out of D in particular. This is for the same reason as we saw in the duty-
of-aid case. There is something special about the public “in your face” char-
acter of exploitation and the failure to do one’s duty to this particular person
that makes this action more problematic than mere failure to contribute to
the opportunities of people generally.
There is a worry with any such set of requirements that they may have the
perverse effect of dissuading the wealthy from employing the worse off.
This could have the perverse effect of making the worse off even worse off
than they would be were they to be taken advantage of. This does seem
like a legitimate worry. It is somewhat analogous to a leveling down
worry.
My response to this invokes the definition of unfair advantage taking,
which includes a violation of a prior duty. The idea is that the wealthy can-
not avoid the duties of contributing their fair shares to the worse off by
avoiding employment relations. They still owe the duties. And they owe
them to everyone. So the fair share they would be required to pay would be
required anyway. In effect, it would determine a baseline. To say that one
does not want to employ very poor persons on the ground that one does not
want to discharge the duty of repair and thereby avoid the charge of unfair
advantage taking is confused. One owes the duty anyway. The question is
how one discharges it.
But this does suggest another puzzle. Could the wealthy pay a sum
towards some development agency and then take full advantage of the
worse-off persons in labor contracts? It isn’t clear this would make any sense
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
economically but in any case it seems problematic. The reason why this is
problematic is because taking advantage of the unequal opportunities of
(say) developing-world workers is in some way worse than merely failing
to discharge the duty of opportunity enhancement. Taking advantage of
unfortunate workers is a kind of public and “in your face” way of treating
persons as less than equal. It is highly expressive of a failure to recognize the
worker as an equal.
Failing to discharge one’s duty more generally, since it is not directed at
any particular person, does not have the same public meaning.
I have argued that wrongful exploitation can only be explained with the help
of a prior duty to a person that is violated, which duty is independent of and
prior to the duty not to exploit. But I think there is, nevertheless, a distinc-
tive duty not to exploit. This is because exploitation introduces a distinctive
dimension to wronging a person. It in some crucial way seems perversely to
make the person a part of the wronging. It makes the wronged person a
participant at least in the sense that the person is acting in a way that makes
exploitation possible. In contrast, think of wrongfully benefitting from your
wrong to another person that is not exploitative. If A kills his parent so as to
receive the inheritance from the parent, A is benefitting from wronging the
parent. But A is not exploiting the parent because the parent’s actions do not
currently contribute to the illicit advantage A gains. In contrast, A’s exploita-
tion of B involves A’s deriving a benefit from B’s activities. B is involved. This
seems to make it a more intimate kind of wrong to B. It seems an even more
public and clear way in which B’s interests are subordinated to A.26
7. CONCLUSION
26
See Snyder (2013).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
tion. I have also tried to show how the account can avoid the problem of
counterproductivity or leveling down that seems to attend many views of
wrongful exploitation. I take this to be a major point in favor of the view.27
Bibliography
Aquinas, Thomas (2003). On Law, Morality and Politics. Ed. Richard J. Regan and
William P. Baumgarth. Indianapolis: Hackett.
Arneson, Richard (1982). “What’s Wrong With Exploitation?” Ethics (91) 2,
pp. 202–27.
Arneson, Richard (1996). Book Review of Exploitation, by Alan Wertheimer, Mind
(110) 439: pp. 888–91.
Bigwood, Rick (2004). Exploitative Contracts. Oxford: Oxford University Press.
Christiano, Thomas (2013). “Equality, Fairness and Agreements,” Journal of Social
Philosophy: Special Issue on New Directions in Egalitarianism, Vol. 44, n. 4 (Winter
2013): pp. 1–22.
Christiano, Thomas and Braynen, Will (2008). “Inequality, Injustice and Leveling
Down,” Ratio, Vol. XXI, n. 4 (December).
Julius, A. J. (2013). “The Possibility of Exchange,” Politics, Philosophy and Economics,
Vol. 12, n. 4 (November).
Marx, Karl (2003). Capital: A Critique of Political Economy in The History of Economic
Thought: A Reader. Ed. Steven Medema and Warren Samuels. London: Routledge.
Mill, John Stuart (2008). Principles of Political Economy. Ed. Jonathan Riley. Oxford:
Oxford University Press.
Pogge, Thomas (2002). World Poverty and Human Rights. London: Polity Press.
Pogge, Thomas (2009). “Testing Our Drugs on the Poor Abroad,” in Exploitation
and Developing Countries. Ed. Jennifer Hawkins and Ezekiel Emanuel. Princeton,
NJ: Princeton University Press.
Roemer, John (1985). “Should Marxists Be Interested in Exploitation?” Philosophy
and Public Affairs, Vol. 14, n. 1 (Winter), pp. 30–65.
Satz, Debra (2010). Why Some Things Should Not Be for Sale: The Moral Limits of
Markets. Oxford: Oxford University Press.
Snyder, J. (2013). “Demeaning Choices,” Politics, Philosophy and Economics, Vol. 12,
n. 4: 313–58.
Steiner, Hillel (1984). “A Liberal Theory of Exploitation,” Ethics Vol. 94, pp. 225–41
Valdman, Mikhail (2009). “A Theory of Wrongful Exploitation,” Philosopher’s
Imprint, Vol. 9, n. 6 (July).
27
I would like to thank the participants in the Oxford Studies in Political Philosophy
Conference in Tucson, October 2013, the participants in the Rocky Mountain Ethics
Conference in Boulder, Colorado, August 2012, and the participants in my seminar on
Exploitation in the Spring of 2013 for helpful comments. In particular, I thank Violeta
Ignieska, Stefan Sciaraffa, David Estlund, Cynthia Stark, Alistair Norcross, Michael
Tooley, David Schmidtz, Houston Smit, Meena Krishnamirthy, Andrew Williams, and
an anonymous referee for OSPP for valuable discussion on the chapter.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
PA RT I V
METHOD
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
11
Value-freeness and Value-neutrality in
the Analysis of Political Concepts
Ian Carter
allow us simply to describe how things are in the world, without presuppos-
ing or implying any ethical judgments, so that power and freedom qualify
as “value-free” or “value-neutral” concepts? Do power and freedom have
negative and positive value in the same ways as those in which, say, injustice
and justice do? If so, it seems that to describe an interpersonal relation as
one of power is necessarily to disapprove of it (at least pro tanto), and to
describe a situation as one of freedom is necessarily to approve of it (at least
pro tanto). Or is it the case that we first describe a particular situation as one
of power or freedom, and then evaluate it ethically? The latter view seems
plausible in light of the fact that a supporter of the American Constitution
might, after all, judge the power exercised by Barack Obama to be ethically
permissible or required, and an Islamic fundamentalist might, after all, judge
the less free society (on a liberal definition of freedom) to be ethically supe-
rior. It matters which of these views is correct, for the answer will affect how
we understand the very nature of phenomena like power and freedom and,
in turn, the ways in which we justify certain evaluations or prescriptions.
In this chapter I shall defend the use of value-free and value-neutral con-
cepts in political philosophy. As I see it, some reasons for working with such
concepts have been overlooked, owing to a failure to make certain relevant
distinctions. This failure has brought about a misleading polarization of the
debate: on the one side, there are those who affirm the possibility and use-
fulness of defining certain political concepts in a way that is wholly detached
from any ethical concerns; on the other side, there are those who deny the
usefulness or even the possibility of a value-free or value-neutral analysis of
any of the concepts that figure in our normative political discourse.
The position I shall defend is more complex, in two senses. First, I shall
propose a distinction between value-freeness and value-neutrality (I use the
slightly strange-sounding noun “value-freeness,” in preference to “value-
freedom,” to avoid confusion with freedom understood as a political value).
Value-freeness concerns the place (or rather, the absence) of evaluative terms
in the definition of a concept, whereas value-neutrality involves a suspen-
sion of judgment about the comparative merits of divergent ethical posi-
tions. Each of these two features is a methodological desideratum in the case
of certain concepts in certain theoretical contexts. Sometimes, it is appro-
priate to adopt an empirical approach to the analysis and definition of a
concept, and this requires a value-free definition. Sometimes, it is appropri-
ate to adopt an ethically non-committal approach, and this requires a degree
of value-neutrality. The two approaches can be combined, but, as we shall
see, they can also come apart.
Second, I shall propose a distinction between value-freeness and value-
neutrality, on the one hand, and the complete detachment of our analysis
from all ethical concerns, on the other. I shall call the latter idea “value-
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
A number of philosophers have defended the view that political concepts fall
into two categories: the evaluative and the non-evaluative. On this view,
freedom and power are prime examples of non-evaluative political concepts.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
Other concepts that have sometimes been said to fall into this category
include equality, democracy, and law. Evaluative concepts, on the other
hand, include justice and desert: unlike in the case of freedom and power,
one cannot say that a certain distribution of resources is just, or that a certain
person deserves a reward, without thereby making a normative ethical claim.
Authors who have taken this dualist or “separationist” view of political con-
cepts have included Felix Oppenheim (1961, 1981), Hillel Steiner (1994),
Matthew Kramer (2003), and Frank Lovett (2010). According to Oppenheim,
for example, we should define social freedom in the following way: agent A
is free to perform some action x if and only if no other agent is rendering A’s
performance of x impossible or punishable (1981, p. 53). Whether or not
some other agent is rendering x impossible or punishable can be ascertained
empirically without recourse to value judgments (as long as we can in turn
plausibly understand the concept of punishability in purely empirical
terms1). Am I free to attend the church down my road, or to vote, or to
leave the country? The answer depends on whether some other agent is
making it impossible for me to do so, or would successfully punish me were
I to do so. Statistical evidence will provide us with empirical (probabilistic)
answers to such questions. Freedom is, therefore, a different kind of concept
from the concepts of justice and desert.
But this view has been hotly contested by those who see all political con-
cepts as necessarily value-laden. According to these authors, who include
William Connolly (1993), Steven Lukes (2005), and Ronald Dworkin
(1986, 2004, 2011), freedom, power, democracy, and law, no less than jus-
tice, desert, and the good life, must be analyzed and defined as part of a
system of ethical values. Each of them is inescapably an ethical concept.
Thus, in the case of freedom, these authors hold that one cannot specify
what a person is free or unfree to do without first specifying what it is valu-
able for him or her to do, and one cannot compare the degree of freedom of
one person with that of another without comparatively evaluating the ends
that those freedoms permit them to realize.
Those in the first camp have sometimes claimed that their non-evaluative
definitions of concepts like freedom or power are neutral between compet-
ing ethical standpoints. This “value-neutrality” is said to be useful because it
provides us with a shared starting point in terms of which to express genu-
ine ethical disagreements. Oppenheim illustrates this point with the follow-
ing example. Imagine that in my opinion freedom is best realized under
1
One person’s punishment might be another person’s reward. However, such evalua-
tions of sanctions can themselves be objects of empirical investigation. Steiner (1994),
Carter (1999), and Kramer (2003), on the other hand, define unfreedom only in terms
of physical impossibility.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
used. They cannot be correctly defined from some perspective wholly out-
side ethics. Ethical and political concepts, including freedom no less than
justice, are “interpretive” concepts, and are in this sense different from the
“criterial” concepts used in the empirical sciences (Dworkin 2004; 2011,
Ch. 8). The concept of a lion can be explicated by setting out empirical
criteria, such as the DNA of lions. The concepts of freedom and justice
cannot be explicated in this way, but are properly interpreted and clarified
by reference to the system of ethical values of which they are a part. For
example, if the ideal of freedom is part of an egalitarian ethical vision of
society (as it is for Dworkin), it needs to be interpreted and defined in terms
of egalitarian values.
I suggest that there are three distinct notions at work in the above debate,
and that these notions can come apart in interesting ways. First, there is the
notion that I call “value-freeness.” This notion can be defined as follows:
Value-freeness: a concept is value-free if its definition is such that the
definiens contains no evaluative terms.
I assume that an evaluative term expresses an evaluative concept. When moral
philosophers refer to an “evaluative concept” they usually mean a concept the
use of which necessarily involves an evaluation, or what I shall here call an
“essentially evaluative concept.” There are, however, many concepts that are
not themselves evaluative in this strict sense but are often used evaluatively.
These are concepts the use of which does not necessarily express an evaluation,
but often has evaluative connotations. I shall call these concepts “non-essen-
tially evaluative.” Hillary Putnam (1981, p. 209) provides the following exam-
ple: “in our culture ‘slobbers his food all over his shirt’ has strong negative
emotive force although the phrase is literally a description.” I shall assume here
that a definition counts as value-free if the definiens is free not merely of essen-
tially evaluative terms but also of non-essentially evaluative terms—that is, of
evaluative terms in the broader sense. More generally, when I refer to “evalu-
ative concepts” without further qualification, I should be taken to refer to
evaluative concepts in this broader sense. In addition, I assume that what
counts as a non-essentially evaluative concept (as opposed to a straight for-
wardly non-evaluative concept) depends on the evaluative stance of the per-
son using it—that is, it depends on whether the concept serves, from its user’s
ethical point of view, to provide evaluations of the world.
When Oppenheim claims that freedom and power should be given
“empirical” definitions, or are “purely descriptive” concepts, he means that
they should be treated as value-free in my sense. Similarly, when Connolly
and Dworkin claim that these concepts are evaluative, they mean that they
are essentially evaluative in my sense and cannot or should not be treated as
value-free in my sense.
Second, there is the notion of value-neutrality, which I define as follows:
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
2
I take the term “value-independence”, and its meaning, from Kramer (forthcoming).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
3
Dworkin’s notions of “austerity” and “neutrality” (1996) bear some resemblance,
respectively, to my notions of “value-freeness” and “value-neutrality.” However, for
Dworkin, austerity and neutrality are themselves metaethical notions that characterize
the position he calls “archimedeanism,” and they come apart only in the case of moral
error theory (which affirms austerity without neutrality).
4
Value-independence seems to be the closest of my three notions to Weberian Wertfreiheit,
even though the latter is normally translated as “value freedom” or “value neutrality.”
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
2. VALUE-FREENESS
5
I use the term “properties,” here, simply to mean those characteristics that we must
recognize objects as having in order to refer to the ways in which they are alike or differ
from one another. This “superficial” usage of the term “property” neither presupposes nor
contradicts realism about ethical properties. A similar point applies where I speak of
freedom or justice as “present in the world.”
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
text is with value-free concepts that are themselves evaluative in the broad
sense. Moreover, even essentially evaluative concepts are used to describe.
The distinction between descriptive and evaluative concepts is therefore
potentially misleading.
This is not to deny that there is an important distinction between descrip-
tion and evaluation. Peter Morriss has illustrated this distinction by consid-
ering the way in which we use the term “persecution.” It would be strange
to say: “A lot of persecuting is going on and all the better for it.” In making
a statement about the presence of persecution in a political situation, I am
automatically taken to be presenting a (prima facie) negative evaluation of
that situation. However, Morriss goes on, if the activities of evaluation and
description were not distinct, I could not be taken to have a reason for my
evaluation. When I say that a particular society contains persecution, what
I am saying is in fact short for: “I condemn this situation (evaluation) and
the reason I condemn it is that it involves persecution (description)”: “The
speaker has to have some descriptive content in mind when using the term
‘persecution’, otherwise he is just saying that he condemns the society for no
reason at all” (Morriss 2002, p. 201).6
Morriss is certainly right to distinguish in this way between the evaluative
and descriptive content of evaluative judgments. However, the distinction
will not suffice to further the cause of value-freeness in the analysis of polit-
ical concepts, for the concepts of freedom and justice each have both a
descriptive and an evaluative function: if a liberal describes a country as
“free,” she implicitly (and prima facie) evaluates the country in a positive
way; if a champion of justice evaluates the outcome of a criminal trial as
“just,” she describes that outcome as, say, something that has come about as
the result of a certain procedure having been followed (or, if the outcome is
considered in isolation from the procedure that produced it, as a result that
reflects the truth about a person’s guilt or innocence). Claims about free-
dom both describe and evaluate; and claims about justice both describe and
evaluate. There is nothing less material about justice than there is about
freedom. It is true that when we ascribe freedom or power to a person or
situation, we are saying something definite about the non-evaluative, empir-
ical world: we are saying that certain material pushings, pullings, influences,
6
Moral philosophers echo this point when they say that a “thick” ethical concept (of
which persecution is presumably an example) can be analyzed into a “thin” evaluative
component and a descriptive component. I shall not make use of the distinction between
“thick” and “thin” evaluative concepts here, as I share Samuel Sheffler’s misgivings about
the clarity of such a dichotomy. Is justice a “thick” or a “thin” evaluative concept?
(Sheffler 1987, p. 417). Moreover, the hierarchical model I present below (in section 2.2)
implies that all evaluative concepts, including supposedly “thin” ones, can have descrip-
tive content.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
7
This is not, of course, “the concept of justice” as Rawls himself defines it at p. 5 of
A Theory of Justice (1999). I return to Rawls’s own definition in section 3.3.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
that ground those referred to in the definiens—in this example, the property
of “being due to a person.” The grounding properties tell us what counts as an
instantiation of the grounded property. The political liberal might say, for
example, that giving each person her due consists in respecting her rights:
respect for rights is what is due to people, it is what counts as people’s due;
justice is grounded in (that is, it exists in virtue of ) respect for people’s rights.
She might then go on to ground rights in the satisfaction of certain funda-
mental interests. And she might then continue in her explication of her con-
ception of justice by arguing that one such fundamental interest consists in
an interest in freedom. This substantive conception of justice would be
advocated by appeal to a so-called “interest theory” of rights together with
a justification of the fundamental interest in freedom: justice is grounded in
respect for rights, is grounded in the satisfaction of interests, is grounded
in the protection or promotion of freedom.
Elsewhere I have defended this ethical construction by appeal to the
“non-specific” or “content-independent” value of freedom (Carter 1995,
1999; cf. van Hees 2000; Kramer 2003). The non-specific (or content-inde-
pendent) value of freedom is the value possessed by freedom independently
of the fact of its consisting in the freedom to do one or another specific
thing. It is the value that freedom has as such. If people have an interest in
freedom as such, there will be contexts in which it is appropriate to increase
or promote their freedom. Their rights protect, among other things, their
interest in having a measure of freedom. And this measure must be ascer-
tained empirically, without reference to the value of the freedom specifically
to do this or that thing (Carter 1999, Part 3).8 This particular theory of
freedom provides just one example of the way in which one might justify
locating a particular property at the bottom of one’s grounding hierarchy of
evaluative properties.
Some evaluative properties, then, are located at the bottom of a grounding
hierarchy of evaluative properties, and are not themselves grounded in eval-
uative properties. Consider our example of a liberal theory of justice as
grounded in a right to freedom. If we continue to descend beyond this hier-
archy of evaluative properties, we shall pass from particular instantiations of
8
Because Dworkin ignores the reasons we have for valuing freedom non-specifically,
he cannot see how freedom could be located at the bottom of the hierarchy without los-
ing its ethical value. And yet he cannot avoid locating freedom, implicitly, at the bottom
of the hierarchy. He therefore distinguishes between mere empirical freedom and the
truly ethical value of “liberty,” which he implicitly locates much higher up in the hierar-
chy (Dworkin 2011, Ch. 17). It is not clear what this moralized concept of liberty adds to
an interpretation of justice based on equality, rights, and empirical freedom. Neither is it
clear how much conceptual significance should be assigned to a linguistic distinction that
can only be made in English.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
9
According to Kramer (2009, p. 211), the thesis that moral properties are “generated
by” (or strongly supervene on) empirical properties is “unexceptional,” but is also “a
profoundly ethical thesis.” If so, that thesis can be understood unproblematically as part
of the overall ethical point of view on which our value-free definition of freedom depends.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
10
Matthew Kramer (forthcoming) summarizes this position effectively, although he
uses the term “value-neutral” where I am here using the term “value-free” (I too have used
the term “value-neutral” in this way in some previous writings).
11
In his critique of value-free definitions of freedom (or what he calls “opportunity”),
Robert Sugden objects that there is no “neutral perspective” in terms of which we can
measure people’s available action, and concludes, skeptically and paradoxically, that
although we value opportunity as such (or as I would put it, although we ascribe non-
specific value to freedom), it is impossible to say how much opportunity (or freedom)
there is in the world (Sugden 2003, p. 802). Presumably, by a “neutral perspective,”
here, Sugden means what I would call a “value-independent” perspective, in which case
the distinction between value-freeness and value-independence ought to save us from
his puzzling conclusion. This is not to deny that the empirical measure I proposed (in
Carter 1999) might require revision.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
3. VALUE-NEUTRALITY
Consider, for example, the dispute about freedom between Isaiah Berlin
(2002) and Gerald MacCallum (1967). According to Berlin, there are at
least two “concepts of freedom”: the negative concept that characterizes the
liberal tradition (roughly speaking, freedom as the absence of humanly
imposed constraints) and the positive concept that characterizes other tra-
ditions such as socialism or communitarianism (freedom as self-mastery or
self-realization). According to MacCallum, on the other hand, here is only
one “concept of freedom”: freedom as a triadic relation between an agent,
constraints, and doings or becomings. Whether one is talking about nega-
tive or positive freedom may depend, according to MacCallum, on whether
the relevant agent is an individual (as in the liberal tradition), or a collectiv-
ity; it may depend on whether one characterizes the relevant “constraints”
more narrowly, so as to include only physical prevention or the use of threats
(as in the liberal tradition), or more broadly, so as to include various kinds
of social pressure, fear, ignorance, or uncontrolled passions; and it may
depend on whether the relevant doings or becomings cover the class of all
possible outcomes (as in the liberal tradition), or only those that the
“authentic” or “rational” agent would choose to realize. People refer to dif-
ferent kinds of “agent,” “constraints,” and “doings/becomings” when talking
about freedom. But, MacCallum says, ultimately they are talking about the
same thing: freedom, understood as a triadic relation.
Since Berlin and MacCallum are theorizing at different levels of abstrac-
tion, they are not really in disagreement but are arguing to different pur-
poses (Ricciardi 2007, pp. 129–30). Berlin’s purpose is to elucidate certain
influential and contrasting ethical visions that make use of different con-
cepts of freedom and to help us decide which of these visions is the most
convincing. His purpose is that of normative theorizing (albeit with a his-
torical dimension). MacCallum’s purpose is metatheoretical: it is not to take
sides in the dispute about whether negative or positive freedom matters
most, but to explain why it is that we call both of them concepts of freedom.
The idea is to point to a basic conceptual structure that effectively captures
and clarifies our shared sense that the advocates of negative and positive
freedom, despite using different concepts, are nevertheless all talking about
“the same thing.” Advocates of positive and negative freedom admittedly
use the word “freedom” to refer to quite different concepts; these different
concepts pick out different properties; and yet, grounded in the disjunction
of these properties is a single abstract property—the property that is picked
out by MacCallum’s concept of freedom. By jointly recognizing the signifi-
cance of this abstract property, the two sides come to see that they have
more in common than a mere use of the same word.
The kind of value-neutrality achieved by MacCallum is quite different
from that envisaged by Oppenheim. Oppenheim fills in the three abstract
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
would have called a “category.” For Berlin, “there exist central features of
our experience that are invariant and omnipresent, or at least much less
variable than the vast variety of its empirical characteristics.” These are the
“basic categories (with their corresponding concepts) in terms of which we
define men” (Berlin 1980, pp. 165–66). Less ambitiously, the aim may be to
defend, clarify or problematize certain abstract notions that are implicit in
a very broad set of normative theories—for example, abstract notions of law
(Hart 1994) or rights (Hohfeld 2001) or desert (Kagan 2012) or “fitting-
ness” (Cupit 1996) or, as we have seen, freedom. Abstract metatheoretical
reflection of this sort has been somewhat neglected in contemporary politi-
cal philosophy. It does not produce policy prescriptions, but it is no less
important for that. Political philosophy is, after all, a branch of philosophy.
12
A mistake that I too have made in earlier writings. For example: Carter (1999),
pp. 15–17.
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
13
Although MacCallum does not use the concept-conception distinction, he does
seem to think that he has clarified the concept of freedom (MacCallum 1967, p. 320).
14
This seems to have been Hart’s view (1994, pp. 246–7).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
4. CONCLUSION
Bibliography
Berlin, Isaiah (1980). Concepts and Categories (Oxford: Oxford University Press).
Berlin, Isaiah (2002). “Two Concepts of Liberty,” in I. Berlin, Liberty (Oxford:
Oxford University Press).
Carter, Ian (1995). “The Independent Value of Freedom,” Ethics, 105: 819–45.
Carter, Ian (1999). A Measure of Freedom (Oxford: Oxford University Press).
Connolly, William E. (1993). The Terms of Political Discourse, third edition (Oxford:
Blackwell).
Correia, Fabrice, and Schnieder, Benjamin (eds) (2012). Metaphysical Grounding.
Understanding the Structure of Reality (Cambridge: Cambridge University Press).
Cupit, Geoffrey (1996). Justice as Fittingness (Oxford: Clarendon Press).
Day, J. P. (1987). Liberty and Justice (London: Croom Helm).
Dworkin, Ronald (1986). Law’s Empire (London: Fontana Press).
Dworkin, Ronald (1996). “Objectivity of Truth: You’d Better Believe It,” Philosophy
and Public Affairs, 25: 87–139.
Dworkin, Ronald (2004). “Hart’s Postscript and the Character of Political
Philosophy,” Oxford Journal of Legal Studies, 24: 1–37.
Dworkin, Ronald (2011). Justice for Hedgehogs (Cambridge, MA: Harvard University
Press).
Hare, R. M. (1984). “Supervenience,” Proceedings of the Aristotelian Society, 58: 1–16.
Hart, H. L. A. (1994). The Concept of Law, second edition (Oxford: Clarendon
Press).
Hees, M. Van (2000). Legal Reductionism and Freedom (Dordrecht: Kluwer).
Hohfeld, Wesley Newcomb (2001). Fundamental Legal Conceptions as Applied in
Judicial Reasoning (Dartmouth: Ashgate).
Jackson, Frank (1998). From Metaphysics to Ethics. A Defence of Conceptual Analysis
(Oxford: Clarendon Press).
Kagan, Shelly (2012). The Geometry of Desert (Oxford: Oxford University Press).
Kovesi, Julius (1967). Moral Notions (London: Routledge and Kegan Paul).
Kramer, Matthew H. (2003). The Quality of Freedom (Oxford: Oxford University
Press).
Kramer, Matthew H. (2009). Moral Realism as a Moral Doctrine (Oxford: Blackwell).
Kramer, Matthew H. (forthcoming). “Conceptual Analysis and Distributive Justice,”
in S. Olsaretti (ed.), The Oxford Handbook of Distributive Justice (Oxford: Oxford
University Press).
Lovett, Frank (2010). A General Theory of Domination and Justice (Oxford: Oxford
University Press).
Lukes, Steven (2005). Power. A Radical View, second edition (London: Macmillan).
MacCallum, Gerald C. Jr (1967). “Negative and Positive Freedom,” The Philosophical
Review, 76: 312–34.
McLaughlin, Brian, and Bennett, Karen (2011). “Supervenience,” in Stanford
Encyclopedia of Philosophy, <http://plato.stanford.edu/entries/supervenience>.
Morriss, Peter (2002). Power. A Philosophical Analysis, second edition (Manchester:
Manchester University Press).
OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
Index
308 Index
Index 309
310 Index
Index 311
312 Index
Index 313
agency effects 47–9, 55–6 Waldron, Jeremy 14, 114, 167, 229
action-guiding advice 49–52 Wall, Steven 89–90
choice and 48 war
causal efficacy of 40–1 collectivist accounts 173
Goldman’s account 40–1 Conditional Force Argument 174–6
sufficient conditions for reductivist account 173
outcome 42 duties of care 183–6
Tuck’s account 38–41 genocidal aggression 180–2
causal responsibility for outcome 43, lesser interests 187–8
43–4, 57 mediated harms 176–8
expressive account 53–5 proliferation problem 188–9
perceived 56–7 rescue cases 182–3
expressive account 37 vital interests 178–9
electoral responsibility 53–5 Welfare as Reason 92–3, 101, 104
moral responsibility 45–9 Wellman, Christopher Heath 154–8
for outcome 54 Wertheimer, Alan 254, 257, 261
pivotal votes 36–7 Williams, Bernard 293
political attitudes and 53–4 Wood, Allen 252
rational choice theory 36–45 wrongful exploitation 5
rational ignorance 37 wrongful subjection 169–71
rationality of 1–2, 37
Vrousalis, Nicholas 266 Zwolinski, Matt 261